summaryrefslogtreecommitdiff
diff options
context:
space:
mode:
-rw-r--r--.gitattributes4
-rw-r--r--LICENSE.txt11
-rw-r--r--README.md2
-rw-r--r--old/66286-0.txt3561
-rw-r--r--old/66286-0.zipbin61443 -> 0 bytes
-rw-r--r--old/66286-h.zipbin2303012 -> 0 bytes
-rw-r--r--old/66286-h/66286-h.htm4724
-rw-r--r--old/66286-h/images/cbl-3.jpgbin10982 -> 0 bytes
-rw-r--r--old/66286-h/images/cover.jpgbin226555 -> 0 bytes
-rw-r--r--old/66286-h/images/i_010a.jpgbin30582 -> 0 bytes
-rw-r--r--old/66286-h/images/i_010b.jpgbin23318 -> 0 bytes
-rw-r--r--old/66286-h/images/i_010c.jpgbin20017 -> 0 bytes
-rw-r--r--old/66286-h/images/i_010d.jpgbin17889 -> 0 bytes
-rw-r--r--old/66286-h/images/i_011a.jpgbin19234 -> 0 bytes
-rw-r--r--old/66286-h/images/i_011b.jpgbin20132 -> 0 bytes
-rw-r--r--old/66286-h/images/i_011c.jpgbin20053 -> 0 bytes
-rw-r--r--old/66286-h/images/i_011d.jpgbin23743 -> 0 bytes
-rw-r--r--old/66286-h/images/i_011e.jpgbin20611 -> 0 bytes
-rw-r--r--old/66286-h/images/i_012a.jpgbin28356 -> 0 bytes
-rw-r--r--old/66286-h/images/i_012b.jpgbin20244 -> 0 bytes
-rw-r--r--old/66286-h/images/i_012c.jpgbin24388 -> 0 bytes
-rw-r--r--old/66286-h/images/i_013a.jpgbin41724 -> 0 bytes
-rw-r--r--old/66286-h/images/i_013b.jpgbin34027 -> 0 bytes
-rw-r--r--old/66286-h/images/i_016.jpgbin49392 -> 0 bytes
-rw-r--r--old/66286-h/images/i_017.jpgbin54630 -> 0 bytes
-rw-r--r--old/66286-h/images/i_019.jpgbin80470 -> 0 bytes
-rw-r--r--old/66286-h/images/i_035.jpgbin94423 -> 0 bytes
-rw-r--r--old/66286-h/images/i_037a.jpgbin82062 -> 0 bytes
-rw-r--r--old/66286-h/images/i_037b.jpgbin58996 -> 0 bytes
-rw-r--r--old/66286-h/images/i_038.jpgbin93414 -> 0 bytes
-rw-r--r--old/66286-h/images/i_039a.jpgbin49025 -> 0 bytes
-rw-r--r--old/66286-h/images/i_039b.jpgbin78468 -> 0 bytes
-rw-r--r--old/66286-h/images/i_041a.jpgbin30046 -> 0 bytes
-rw-r--r--old/66286-h/images/i_041b.jpgbin42333 -> 0 bytes
-rw-r--r--old/66286-h/images/i_041c.jpgbin94146 -> 0 bytes
-rw-r--r--old/66286-h/images/i_055.jpgbin51735 -> 0 bytes
-rw-r--r--old/66286-h/images/i_056a.jpgbin39276 -> 0 bytes
-rw-r--r--old/66286-h/images/i_056b.jpgbin53056 -> 0 bytes
-rw-r--r--old/66286-h/images/i_057a.jpgbin44935 -> 0 bytes
-rw-r--r--old/66286-h/images/i_057b.jpgbin55053 -> 0 bytes
-rw-r--r--old/66286-h/images/i_057c.jpgbin58928 -> 0 bytes
-rw-r--r--old/66286-h/images/i_057d.jpgbin59673 -> 0 bytes
-rw-r--r--old/66286-h/images/i_057e.jpgbin50054 -> 0 bytes
-rw-r--r--old/66286-h/images/i_060.jpgbin74827 -> 0 bytes
-rw-r--r--old/66286-h/images/i_061a.jpgbin36060 -> 0 bytes
-rw-r--r--old/66286-h/images/i_061b.jpgbin40711 -> 0 bytes
-rw-r--r--old/66286-h/images/i_062.jpgbin58045 -> 0 bytes
-rw-r--r--old/66286-h/images/i_064.jpgbin48316 -> 0 bytes
-rw-r--r--old/66286-h/images/i_065.jpgbin55897 -> 0 bytes
-rw-r--r--old/66286-h/images/i_066.jpgbin61058 -> 0 bytes
-rw-r--r--old/66286-h/images/i_067a.jpgbin100165 -> 0 bytes
-rw-r--r--old/66286-h/images/i_067b.jpgbin100835 -> 0 bytes
-rw-r--r--old/66286-h/images/i_070a.jpgbin52510 -> 0 bytes
-rw-r--r--old/66286-h/images/i_070b.jpgbin48009 -> 0 bytes
-rw-r--r--old/66286-h/images/i_081.jpgbin58269 -> 0 bytes
55 files changed, 17 insertions, 8285 deletions
diff --git a/.gitattributes b/.gitattributes
new file mode 100644
index 0000000..d7b82bc
--- /dev/null
+++ b/.gitattributes
@@ -0,0 +1,4 @@
+*.txt text eol=lf
+*.htm text eol=lf
+*.html text eol=lf
+*.md text eol=lf
diff --git a/LICENSE.txt b/LICENSE.txt
new file mode 100644
index 0000000..6312041
--- /dev/null
+++ b/LICENSE.txt
@@ -0,0 +1,11 @@
+This eBook, including all associated images, markup, improvements,
+metadata, and any other content or labor, has been confirmed to be
+in the PUBLIC DOMAIN IN THE UNITED STATES.
+
+Procedures for determining public domain status are described in
+the "Copyright How-To" at https://www.gutenberg.org.
+
+No investigation has been made concerning possible copyrights in
+jurisdictions other than the United States. Anyone seeking to utilize
+this eBook outside of the United States should confirm copyright
+status under the laws that apply to them.
diff --git a/README.md b/README.md
new file mode 100644
index 0000000..d3a5591
--- /dev/null
+++ b/README.md
@@ -0,0 +1,2 @@
+Project Gutenberg (https://www.gutenberg.org) public repository for
+eBook #66286 (https://www.gutenberg.org/ebooks/66286)
diff --git a/old/66286-0.txt b/old/66286-0.txt
deleted file mode 100644
index 9c34c3b..0000000
--- a/old/66286-0.txt
+++ /dev/null
@@ -1,3561 +0,0 @@
-The Project Gutenberg eBook of Self-Organizing Systems, 1963, by Various
-
-This eBook is for the use of anyone anywhere in the United States and
-most other parts of the world at no cost and with almost no restrictions
-whatsoever. You may copy it, give it away or re-use it under the terms
-of the Project Gutenberg License included with this eBook or online at
-www.gutenberg.org. If you are not located in the United States, you
-will have to check the laws of the country where you are located before
-using this eBook.
-
-Title: Self-Organizing Systems, 1963
-
-Author: Various
-
-Editor: James Emmett Garvey
-
-Release Date: September 13, 2021 [eBook #66286]
-
-Language: English
-
-Character set encoding: UTF-8
-
-Produced by: Mark C. Orton and the Online Distributed Proofreading Team at
- https://www.pgdp.net
-
-*** START OF THE PROJECT GUTENBERG EBOOK SELF-ORGANIZING SYSTEMS, 1963 ***
-
-
-
-
-Transcriber’s Notes:
-
- Underscores “_” before and after a word or phrase indicate _italics_
- in the original text.
- Equal signs “=” before and after a word or phrase indicate =bold= in
- the original text.
- A single underscore after a symbol indicates a subscript.
- Carat symbol “^” designates a superscript.
- A bold arrow in front of a letter indicates a vector,
- e.g. ⮕M means “the vector M”.
- Small capitals have been converted to SOLID capitals.
- Illustrations have been moved so they do not break up paragraphs.
- Typographical and punctuation errors have been silently corrected.
-
-
-
-
- SELF-ORGANIZING
- SYSTEMS
- 1963
-
- =Edited By=
-
- =JAMES EMMETT GARVEY=
- _Office of Naval Research
- Pasadena, California_
-
- =ACR-96=
-
- =OFFICE OF NAVAL RESEARCH
- DEPARTMENT OF THE NAVY
- WASHINGTON, D.C.=
-
- For sale by the Superintendent of Documents.
- U.S. Government Printing Office
- Washington, D.C., 20402—Price $1.50
-
-
-
-
-CONTENTS
-
-
- Foreword iv
-
- The Ionic Hypothesis and Neuron Models 1
- —E. R. Lewis
-
- Fields and Waves in Excitable Cellular Structures 19
- —R. M. Stewart
-
- Multi-Layer Learning Networks 37
- —R. A. Stafford
-
- Adaptive Detection of Unknown Binary Waveforms 46
- —J. J. Spilker, Jr.
-
- Conceptual Design of Self-Organizing Machines 52
- —P. A. Kleyn
-
- A Topological Foundation for Self-Organization 65
- —R. I. Ścibor-Marchocki
-
- On Functional Neuron Modeling 71
- —C. E. Hendrix
-
- Selection of Parameters for Neural Net Simulations 76
- —R. K. Overton
-
- Index of Invited Participants 77
-
-
-
-
-FOREWORD
-
-
-The papers appearing in this volume were presented at a Symposium
-on Self-Organizing Systems, which was sponsored by the Office of
-Naval Research and held at the California Institute of Technology,
-Pasadena, California, on 14 November 1963. The Symposium was organized
-with the aim of providing a critical forum for the presentation and
-discussion of contemporary significant research efforts, with the
-emphasis on relatively uncommon approaches and methods in an early
-state of development. This aim and nature dictated that the Symposium
-be in effect a Working Group, with numerically limited invitational
-participation.
-
-The papers which were presented and discussed did in fact serve
-to introduce several relatively unknown approaches; some of the
-speakers were promising young scientists, others had become known for
-contributions in different fields and were as yet unrecognized for
-their recent work in self-organization. In addition, the papers as a
-collection provided a particularly broad, cross-disciplinary spectrum
-of investigations which possessed intrinsic value as a portrayal of
-the bases upon which this new discipline rests. Accordingly, it became
-obvious in retrospect that the information presented and discussed at
-the Symposium was of considerable interest—and should thus receive
-commensurate dissemination—to a much broader group of scientists and
-engineers than those who were able to participate directly in the
-meeting itself. This volume is the result of that observation; as an
-edited collection of the papers presented at the Symposium, it forms
-the Proceedings thereof. If it provides a useful reference for present
-and future investigators, as well as documenting the source of several
-new approaches, it will have fulfilled its intended purpose well.
-
-A Symposium which takes the nature of a Working Group depends for its
-utility especially upon effective commentary and critical analysis,
-and we commend all the participants for their contributions in this
-regard. It is appropriate, further, to acknowledge the contributions
-to the success of the Symposium made by the following: The California
-Institute of Technology for volunteering to act as host and for
-numerous supporting services; Professor Gilbert D. McCann, Director
-of the Willis Booth Computing Center at the California Institute of
-Technology, and the members of the technical and secretarial staffs
-of the Computing Center, who assumed the responsibility of acting as
-the immediate representatives of the Institute; the members of the
-Program Committee, who organized and led the separate sessions—Harold
-Hamilton of General Precision, Joseph Hawkins of Ford Motor Company,
-Robert Stewart of Space-General, Peter Kleyn of Northrop, and Professor
-McCann; members of the Technical Information Division of the Naval
-Research Laboratory, who published these Proceedings; and especially
-the authors of the papers, which comprised the heart of the Symposium
-and subsequently formed this volume. To all of these the sponsors wish
-to express their very sincere appreciation.
-
- JAMES EMMETT GARVEY
- _Office of Naval Research Branch Office
- Pasadena, California_
-
- MARGO A. SASS
- _Office of Naval Research
- Washington, D.C._
-
-
-
-
-The Ionic Hypothesis and Neuron Models
-
-
- E. R. LEWIS
-
- _Librascope Group, General Precision, Inc.
- Research and Systems Center
- Glendale, California_
-
- The measurements of Hodgkin and Huxley were aimed at
- revealing the mechanism of generation and propagation
- of the all-or-none spike. Their results led to the
- Modern Ionic Hypothesis. Since the publication of
- their papers in 1952, advanced techniques with
- microelectrodes have led to the discovery of many
- modes of subthreshold activity not only in the axon
- but also in the somata and dendrites of neurons. This
- activity includes synaptic potentials, local response
- potentials, and pacemaker potentials.
-
- We considered the question, “Can this activity also
- be explained in terms of the Hodgkin-Huxley Model?”
- To seek an answer, we have constructed an electronic
- analog based on the ionic hypothesis and designed
- around the data of Hodgkin and Huxley. Synaptic
- inputs were simulated by simple first-order or
- second-order networks connected directly to simulated
- conductances (potassium or sodium). The analog has,
- with slight parameter adjustments, produced all modes
- of threshold and subthreshold activity.
-
-
-INTRODUCTION
-
-In recent years physiologists have become quite adept at probing
-into neurons with intracellular microelectrodes. They are now able,
-in fact, to measure (a) the voltage change across the postsynaptic
-membrane elicited by a single presynaptic impulse (see, for examples,
-references 1 and 2) and (b) the voltage-current characteristics
-across a localized region of the nerve cell membrane (3), (4), (5),
-(6). With microelectrodes, physiologists have been able to examine
-not only the all-or-none spike generating and propagating properties
-of axons but also the electrical properties of somatic and dendritic
-structures in individual neurons. The resulting observations have
-led many physiologists to believe that the individual nerve cell
-is a potentially complex information-processing system far removed
-from the simple two-state device envisioned by many early modelers.
-This new concept of the neuron is well summarized by Bullock in his
-1959 _Science_ article (10). In the light of recent physiological
-literature, one cannot justifiably omit the diverse forms of somatic
-and dendritic behavior when assessing the information-processing
-capabilities of single neurons. This is true regardless of the means of
-assessment—whether one uses mathematical idealizations, electrochemical
-models, or electronic analogs. We have been interested specifically in
-electronic analogs of the neuron; and in view of the widely diversified
-behavior which we must simulate, our first goal has been to find a
-unifying concept about which to design our analogs. We believe we have
-found such a concept in the Modern Ionic Hypothesis, and in this paper
-we will discuss an electronic analog of the neuron which was based on
-this hypothesis and which simulated not only the properties of the
-axon but also the various subthreshold properties of the somata and
-dendrites of neurons.
-
-We begin with a brief summary of the various types of subthreshold
-activity which have been observed in the somatic and dendritic
-structures of neurons. This is followed by a brief discussion of the
-Hodgkin-Huxley data and of the Modern Ionic Hypothesis. An electronic
-analog based on the Hodgkin-Huxley data is then introduced, and we show
-how this analog can be used to provide all of the various types of
-somatic and dendritic activity.
-
-
-SUBTHRESHOLD ELECTRICAL ACTIVITY IN NEURONS
-
-In studying the recent literature in neurophysiology, one is
-immediately struck by the diversity in form of both elicited and
-spontaneous electrical activity in the single nerve cell. This applies
-not only to the temporal patterns of all-or-none action potentials
-but also to the graded somatic and dendritic potentials. The synaptic
-membrane of a neuron, for example, is often found to be electrically
-inexcitable and thus incapable of producing an action potential; yet
-the graded, synaptically induced potentials show an amazing diversity
-in form. In response to a presynaptic impulse, the postsynaptic
-membrane may become hyperpolarized (inhibitory postsynaptic potential),
-depolarized (excitatory postsynaptic potential), or remain at the
-resting potential but with an increased permeability to certain ions
-(a form of inhibition). The form of the postsynaptic potential in
-response to an isolated presynaptic spike may vary from synapse to
-synapse in several ways, as shown in Figure 1. Following a presynaptic
-spike, the postsynaptic potential typically rises with some delay to
-a peak value and then falls back toward the equilibrium or resting
-potential. Three potentially important factors are the delay time
-(synaptic delay), the peak amplitude (spatial weighting of synapse),
-and the rate of fall toward the equilibrium potential (temporal
-weighting of synapse). The responses of a synapse to individual spikes
-in a volley may be progressively enhanced (facilitation), diminished
-(antifacilitation), or neither (1), (2), (7), (8). Facilitation may be
-in the form of progressively increased peak amplitude, or in the form
-of progressively decreased rate of fall (see Figure 2). The time course
-and magnitude of facilitation or antifacilitation may very well be
-important synaptic parameters. In addition, the postsynaptic membrane
-sometimes exhibits excitatory or inhibitory aftereffects (or both) on
-cessation of a volley of presynaptic spikes (2), (7); and the time
-course and magnitude of the aftereffects may be important parameters.
-Clearly, even if one considers the synaptic potentials alone, he is
-faced with an impressive variety of responses. Examples of the various
-types of postsynaptic responses may be found in the literature, but for
-purposes of the present discussion the idealized wave forms in Figure 2
-will demonstrate the diversity of electrical behavior with which one is
-faced.
-
-[Illustration: A. EXCITATORY POSTSYNAPTIC POTENTIAL FROM APLYSIA (SEE
-REF. 2)]
-
-[Illustration: B. EXCITATORY POSTSYNAPTIC POTENTIAL FROM PANULIRUS (SEE
-REF. 1)]
-
-[Illustration: C. EXCITATORY POSTSYNAPTIC POTENTIAL FROM MAMMALIAN
-MOTONEURONE (SEE REF. 24)]
-
-[Illustration: D. PRESYNAPTIC SPIKE
-
-Figure 1—Excitatory postsynaptic potentials in response to a single
-presynaptic spike]
-
-[Illustration: A. TEMPORAL FACILITATION]
-
-[Illustration: B. AMPLITUDE FACILITATION]
-
-[Illustration: C. ANTIFACILITATION]
-
-[Illustration: D. INHIBITORY POSTSYNAPTIC POTENTIALS EXHIBITING
-BIPHASIC REBOUND]
-
-[Illustration: E. PRESYNAPTIC SPIKE BURST
-
-Figure 2—Idealized postsynaptic potentials]
-
-In addition to synaptically induced potentials, low-frequency,
-spontaneous potential fluctuations have been observed in many neurons
-(2), (7), (9), (10), (11). These fluctuations, generally referred to
-as pacemaker potentials, are usually rhythmic and may be undulatory
-or more nearly saw-toothed in form. The depolarizing phase may be
-accompanied by a spike, a volley of spikes, or no spikes at all.
-Pacemaker frequencies have been noted from ten or more cycles per
-second down to one cycle every ten seconds or more. Some idealized
-pacemaker wave forms are shown in Figure 3.
-
-[Illustration: A. PERIODIC BURSTS]
-
-[Illustration: B. PACEMAKER POTENTIALS WITHOUT SPIKES]
-
-[Illustration: C. PACEMAKER POTENTIALS WITH SINGLE SPIKES ON
-DEPOLARIZING PHASE
-
-Figure 3—Idealized pacemaker potentials]
-
-[Illustration: A. FORM OF VOLTAGE STIMULI AND RESULTING MEMBRANE
-POTENTIAL CHANGES.]
-
-[Illustration: B. RESPONSE CURVE OF TYPICAL GRADED RESPONSE REGION.
-
-Figure 4—Graded response]
-
-Bullock (7), (10), (12), (13) has demonstrated the existence of
-a third type of subthreshold response, which he calls the graded
-response. While the postsynaptic membrane is quite often electrically
-inexcitable, other regions of the somatic and dendritic membranes
-appear to be moderately excitable. It is in these regions that Bullock
-observes the graded response. If one applies a series of pulsed voltage
-stimuli to the graded-response region, the observed responses would be
-similar to those shown in Figure 4A. Plotting the peak response voltage
-as a function of the stimulus voltage would result in a curve similar
-to that in Figure 4B (see Ref. 3, page 4). For small values of input
-voltage, the response curve is linear; the membrane is passive. As the
-stimulus voltage is increased, however, the response becomes more and
-more disproportionate. The membrane is actively amplifying the stimulus
-potential. At even higher values of stimulus potential, the system
-becomes regenerative; and a full action potential results. The peak
-amplitude of the response depends on the duration of the stimulus as
-well as on the amplitude. It also depends on the rate of application of
-the stimulus voltage. If the stimulus potential is a voltage ramp, for
-example, the response will depend on the slope of the ramp. If the rate
-of rise is sufficiently low, the membrane will respond in a passive
-manner to voltages much greater than the spike threshold for suddenly
-applied voltages. In other words, the graded-response regions appear to
-accommodate to slowly varying potentials.
-
-In terms of functional operation, we can think of the synapse as a
-transducer. The input to this transducer is a spike or series of spikes
-in the presynaptic axon. The output is an accumulative, long-lasting
-potential which in some way (perhaps not uniquely) represents the
-pattern of presynaptic spikes. The pacemaker appears to perform the
-function of a clock, producing periodic spikes or spike bursts or
-producing periodic changes in the over-all excitability of the neuron.
-The graded-response regions appear to act as nonlinear amplifiers and,
-occasionally, spike initiators. The net result of this electrical
-activity is transformed into a series of spikes which originate at
-spike initiation sites and are propagated along axons to other neurons.
-The electrical activity in the neuron described above is summarized in
-the following outline (taken in part from Bullock (7)):
-
- 1. Synaptic Potentials
- a. Excitatory or inhibitory
- b. Facilitated, antifacilitated, or neither
- c. With excitatory aftereffect, inhibitory aftereffect,
- neither, or both
-
- 2. Pacemaker Potentials
- a. Relaxation type, undulatory type, or none at all
- b. Producing single spike, spike burst, or no spikes
- c. Rhythmic or sporadic
-
- 3. Graded Response (rate sensitive)
-
- 4. Spike Initiation
-
-
-THE MODERN IONIC HYPOTHESIS
-
-Hodgkin, Huxley, and Katz (3) and Hodgkin and Huxley (14), (15), (16),
-in 1952, published a series of papers describing detailed measurements
-of voltage, current, and time relationships in the giant axon of the
-squid (_Loligo_). Hodgkin and Huxley (17) consolidated and formalized
-these data into a set of simultaneous differential equations describing
-the hypothetical time course of events during spike generation and
-propagation. The hypothetical system which these equations describe is
-the basis of the Modern Ionic Hypothesis.
-
-The system proposed by Hodgkin and Huxley is basically one of dynamic
-opposition of ionic fluxes across the axon membrane. The membrane
-itself forms the boundary between two liquid phases—the intracellular
-fluid and the extracellular fluid. The intracellular fluid is rich in
-potassium ions and immobile organic anions, while the extracellular
-fluid contains an abundance of sodium ions and chloride ions. The
-membrane is slightly permeable to the potassium, sodium, and chloride
-ions; so these ions tend to diffuse across the membrane. When the
-axon is inactive (not propagating a spike), the membrane is much more
-permeable to chloride and potassium ions than it is to sodium ions.
-In this state, in fact, sodium ions are actively transported from the
-inside of the membrane to the outside at a rate just sufficient to
-balance the inward leakage. The relative sodium ion concentrations on
-both sides of the membrane are thus fixed by the active transport rate,
-and the net sodium flux across the membrane is effectively zero. The
-potassium ions, on the other hand, tend to move out of the cell; while
-chloride ions tend to move into it. The inside of the cell thus becomes
-negative with respect to the outside. When the potential across the
-membrane is sufficient to balance the inward diffusion of chloride with
-an equal outward drift, and the outward diffusion of potassium with an
-inward drift (and possibly an inward active exchange), equilibrium is
-established. The equilibrium potential is normally in the range of 60
-to 65 millivolts.
-
-The resting neural membrane is thus polarized, with the inside
-approximately 60 millivolts negative with respect to the outside.
-Most of the Hodgkin-Huxley data is based on measurements of the
-transmembrane current in response to an imposed stepwise reduction
-(depolarization) of membrane potential. By varying the external
-ion concentrations, Hodgkin and Huxley were able to resolve the
-transmembrane current into two “active” components, the potassium
-ion current and the sodium ion current. They found that while the
-membrane permeabilities to chloride and most other inorganic ions
-were relatively constant, the permeabilities to both potassium and
-sodium were strongly dependent on membrane potential. In response to a
-suddenly applied (step) depolarization, the sodium permeability rises
-rapidly to a peak and then declines exponentially to a steady value.
-The potassium permeability, on the other hand, rises with considerable
-delay to a value which is maintained as long as the membrane remains
-depolarized. The magnitudes of both the potassium and the sodium
-permeabilities increase monotonically with increasing depolarization.
-A small imposed depolarization will result in an immediately
-increased sodium permeability. The resulting increased influx of
-sodium ions results in further depolarization; and the process
-becomes regenerative, producing the all-or-none action potential.
-At the peak of the action potential, the sodium conductance begins
-to decline, while the delayed potassium conductance is increasing.
-Recovery is brought about by an efflux of potassium ions, and both
-ionic permeabilities fall rapidly as the membrane is repolarized.
-The potassium permeability, however, falls less rapidly than that of
-sodium. This is basically the explanation of the all-or-none spike
-according to the Modern Ionic Hypothesis.
-
-[Illustration: Figure 5—Hodgkin-Huxley representation of small area of
-axon membrane]
-
-[Illustration: Figure 6—Typical responses of sodium conductance and
-potassium conductance to imposed step depolarization]
-
-By defining the net driving force on any given ion species as the
-difference between the membrane potential and the equilibrium potential
-for that ion and describing permeability changes in terms of equivalent
-electrical conductance changes, Hodgkin and Huxley reduced the ionic
-model to the electrical equivalent in Figure 5. The important dynamic
-variables in this equivalent network are the sodium conductance
-(G{Na}) and the potassium conductance (G{K}). The change in the sodium
-conductance in response to a step depolarization is shown in Figure 6B.
-This change can be characterized by seven voltage dependent parameters:
-
- 1. Delay time—generally much less than 1 msec
- 2. Rise time—1 msec or less
- 3. Magnitude of peak conductance—increases
- monotonically with increasing depolarization
- 4. Inactivation time constant—decreases
- monotonically with increasing depolarization.
- 5. Time constant of recovery from
- inactivation—incomplete data
- 6. Magnitude of steady-state conductance—increases
- monotonically with increasing depolarization
- 7. Fall time on sudden repolarization—less than 1 msec.
-
-Figure 6B shows the potassium conductance change in response to
-an imposed step depolarization. Four parameters are sufficient to
-characterize this response:
-
- 1. Delay time—decreases monotonically with
- increasing depolarization
- 2. Rise time—decreases monotonically with increasing
- depolarization
- 3. Magnitude of steady-state conductance—increases
- monotonically with increasing depolarization
- 4. Fall time on sudden repolarization—8 msec
- or more, decreases slightly with increasing
- depolarization.
-
-In addition to the aforementioned parameters, the transient portion of
-the sodium conductance appears to exhibit an accommodation to slowly
-varying membrane potentials. The time constants of accommodation appear
-to be those of inactivation or recovery from inactivation—depending on
-the direction of change in the membrane potential (18). The remaining
-elements in the Hodgkin-Huxley model are constant and are listed below:
-
- 1. Potassium potential—80 to 85 mv (inside negative)
- 2. Sodium potential—45 to 50 mv (inside positive)
- 3. Leakage potential—38 to 43 mv (inside negative)
- 4. Leakage conductance—approx. 0.23 millimhos/cm²
- 5. Membrane capacitance—approx. 1 μf/cm²
- 6. Resting potential—60 to 65 mv
- 7. Spike amplitude—approx. 100 mv
-
-
-ELECTRONIC SIMULATION OF THE HODGKIN-HUXLEY MODEL
-
-[Illustration: Figure 7—System diagram for electronic simulation of the
-Hodgkin-Huxley model]
-
-Given a suitable means of generating the conductance functions,
-G_{Na}(v,t) and G_{K}(v,t), one can readily stimulate the essential
-aspects of the Modern Ionic Hypothesis. If we wish to do this
-electronically, we have two problems. First, we must synthesize
-a network whose input is the membrane potential and whose output
-is a voltage or current proportional to the desired conductance
-function. Second, we must transform the output from a voltage or
-current to an effective electronic conductance. The former implies
-the need for nonlinear, active filters, while the latter implies
-the need for multipliers. The basic block diagram is shown in
-Figure 7. Several distinct realizations of this system have been
-developed in our laboratory, and in each case the results were the
-same. With parameters adjusted to closely match the data of Hodgkin
-and Huxley, the electronic model exhibits all of the important
-properties of the axon. It produces spikes of 1 to 2 msec duration
-with a threshold of approximately 5% to 10% of the spike amplitude.
-The applied stimulus is generally followed by a prepotential, then
-an active rise of less than 1 msec, followed by an active recovery.
-The after-depolarization generally lasts several msec, followed by
-a prolonged after-hyperpolarization. The model exhibits the typical
-strength-duration curve, with rheobase of 5% to 10% of the spike
-amplitude. For sufficiently prolonged sodium inactivation (long time
-constant of recovery from inactivation), the model also exhibits an
-effect identical to classical Wedensky inhibition (18). Thus, as would
-be expected, the electronic model simulates very well the electrical
-properties of the axon.
-
-In addition to the axon properties, however, the electronic model is
-able to reproduce all of the somatic and dendritic activity outlined
-in the section on subthreshold activity. Simulation of the pacemaker
-and graded-response potentials is accomplished without additional
-circuitry. In the case of synaptically induced potentials, however,
-auxiliary networks are required. These networks provide additive terms
-to the variable conductances in accordance with current notions on
-synaptic transmission (19). Two types of networks have been used. In
-both, the inputs are simulated presynaptic spikes, and in both the
-outputs are the resulting simulated chemical transmitter concentration.
-In both, the transmitter substance was assumed to be injected at a
-constant rate during a presynaptic spike and subsequently inactivated
-in the presence of an enzyme. One network simulates a first-order
-chemical reaction, where the enzyme concentration is effectively
-constant. The other simulates a second-order chemical reaction,
-where the enzyme concentration is assumed to be reduced during the
-inactivation process. For simulation of an excitatory synapse, the
-output of the auxiliary network is added directly to G_{Na} in the
-electronic model. For inhibition, it is added to G_{K}. With the
-parameters of the electronic membrane model set at the values measured
-by Hodgkin and Huxley, we have attempted to simulate synaptic activity
-with the aid of the two types of auxiliary networks. In the case of
-the simulated first-order reaction, the excitatory synapse exhibits
-facilitation, antifacilitation, or neither—depending on the setting
-of a single parameter, the transmitter inactivation rate (_i.e._,
-the effective enzyme concentration). This parameter would appear,
-in passing, to be one of the most probable synaptic variables. In
-this case, the mechanisms for facilitation and antifacilitation are
-contained in the simulated postsynaptic membrane. Facilitation is due
-to the nonlinear dependence of G_{Na} on membrane potential, while
-antifacilitation is due to inactivation of G_{Na}. The occurrence
-of one form of response or the other is determined by the relative
-importance of the two mechanisms (18). Grundfest (20) has mentioned
-both of these mechanisms as potentially facilitory and antifacilitory,
-respectively. The simulated inhibitory synapse with the first order
-input is capable of facilitation (18), but no antifacilitation has been
-observed. Again, the presence or absence of facilitation is determined
-by the inactivation rate.
-
-With the simulated second-order reaction, both excitatory and
-inhibitory synapses exhibit facilitation. In this case, two facilitory
-mechanisms are present—one in the postsynaptic membrane and one in the
-nonconstant transmitter inactivation reaction. The active membrane
-currents can, in fact, be removed; and this system will still exhibit
-facilitation. With the second-order auxiliary network, the presence
-of excitatory facilitation, antifacilitation, or neither depends
-on the initial, or resting, transmitter inactivation rate. The
-synaptic behavior also depends parametrically on the simulated enzyme
-reactivation rate. Inhibitory antifacilitation can be introduced with
-either type of auxiliary network by limiting the simulated presynaptic
-transmitter supply.
-
-Certain classes of aftereffects are inherent in the mechanisms of the
-Ionic Hypothesis. In the electronic model, aftereffects are observed
-following presynaptic volleys with either type of auxiliary network.
-Following a volley of spikes into the simulated excitatory synapse,
-for example, rebound hyperpolarization may or may not occur depending
-on the simulated transmitter inactivation rate. If the inactivation
-rate is sufficiently high, rebound will occur. This rebound can be
-monophasic (inhibitory phase only) or polyphasic (successive cycles
-of excitation and inhibition). Following a volley of spikes into the
-simulated inhibitory synapse, rebound depolarization may or may not
-occur depending on the simulated transmitter inactivation rate. This
-rebound can also be monophasic or polyphasic. Sustained postexcitatory
-depolarization and sustained postinhibitory hyperpolarization (2) have
-been achieved in the model by making the transmitter inactivation rate
-sufficiently low.
-
-The general forms of the postsynaptic potentials simulated with
-the electronic model are strikingly similar to those published in
-the literature for real neurons. The first-order auxiliary network
-produces facilitation of a form almost identical to that shown by Otani
-and Bullock (8) while the second-order auxiliary network produces
-facilitation of the type shown by Chalazonitis and Arvanitake (2).
-The excitatory antifacilitation is almost identical to that shown by
-Hagiwara and Bullock (1) in both form and dependence on presynaptic
-spike frequency. In every case, the synaptic behavior is determined
-by the effective rate of transmitter inactivation, which in real
-neurons would presumably be directly proportional to the effective
-concentration of inactivating enzyme at the synapse.
-
-Pacemaker potentials are easily simulated with the electronic model
-without the use of auxiliary networks. This is achieved either by
-inserting a large, variable shunt resistor across the simulated
-membrane (see Figure 5) or by allowing a small sodium current leakage
-at the resting potential. With the remaining parameters of the
-model set as close as possible to the values determined by Hodgkin
-and Huxley, the leakage current induces low-frequency, spontaneous
-spiking. The spike frequency increases monotonically with increasing
-leakage current. In addition, if the sodium conductance inactivation
-is allowed to accumulate over several spikes, periodic spike pairs
-and spike bursts will result. Subthreshold pacemaker potentials have
-also been observed in the model, but with parameter values set close
-to the Hodgkin-Huxley data these are generally higher in frequency
-than pacemaker potentials in real neurons. It is interesting that
-a pacemaker mode may exist in the absence of the simulated sodium
-conductance. It is a very high-frequency mode (50 cps or more)
-and results from the alternating dominance of potassium current
-and chloride (or leakage ion) current in determining the membrane
-potential. The significance of this mode cannot be assessed until
-better data is available for the potassium conductance at low levels
-of depolarization in real neurons. In general, as far as the model is
-concerned, pacemaker potentials are possible because the potassium
-conductance is delayed in both its rise with depolarization and its
-fall with repolarization.
-
-Rate sensitive graded response has also been observed in the electronic
-model. The rate sensitivity—or accommodation—is due to the sodium
-conductance inactivation. The response of the model to an imposed ramp
-depolarization was discussed in Reference 18. At this time, several
-alternative model parameters could be altered to bring about reduced
-electrical excitability. None of the parameter changes was very
-satisfying, however, because none of them was in any way justified by
-physiological data. We have since found that the membrane capacitance,
-a plausible parameter in view of recent physiological findings, can
-completely determine the electrical excitability. Thus, with the
-capacitance determined by Hodgkin and Huxley (1 microfarad per cm²),
-the model exhibits excitability characteristic of the axon. As the
-capacitance is increased, the model becomes less excitable until, with
-10 or 12 μμf, it is effectively inexcitable. Thus, with an increased
-capacitance—but with all the remaining parameters set as close as
-possible to the Hodgkin-Huxley values—the electronic model exhibits the
-characteristics of Bullock’s graded-response regions.
-
-Whether membrane capacitance is the determining factor in real neurons
-is, of course, a matter of speculation. Quite a controversy is raging
-over membrane capacity measurements (see Rall (21)), but the evidence
-indicates that the capacity in the soma is considerably greater than
-that in the axon (6), (22).
-
-It should be added that increasing the capacitance until the membrane
-model becomes inexcitable has little effect on the variety of available
-simulated synaptic responses. Facilitation, antifacilitation, and
-rebound are still present and still depend on the transmitter
-inactivation rate. Thus, in the model, we can have a truly inexcitable
-membrane which nevertheless utilizes the active membrane conductances
-to provide facilitation or antifacilitation, and rebound. The simulated
-subthreshold pacemaker potentials are much more realistic with the
-increased capacitance, being lower in frequency and more natural in
-form.
-
-In one case, the electronic model predicted behavior which was
-subsequently reported in real neurons. This was in respect to the
-interaction of synaptic potentials and pacemaker potential. It was
-noted in early experiments that when the model was set in a pacemaker
-mode, and periodic spikes were applied to the simulated inhibitory
-synapse, the pacemaker frequency could be modified; and, in fact,
-it would tend to lock on to the stimulus frequency. This produced
-a paradoxical effect whereby the frequency of spontaneous spikes
-was actually increased by increasing the frequency of inhibitory
-synaptic stimuli. At very low stimulus frequencies, the spontaneous
-pacemaker frequency was not appreciably perturbed. As the stimulus
-frequency was increased, and approached the basic pacemaker frequency,
-the latter tended to lock on and follow further increases in the
-stimulus frequency. When the stimulus frequency became too high for
-the pacemaker to follow, the latter decreased abruptly in frequency
-and locked on to the first subharmonic. As the stimulus frequency was
-further increased, the pacemaker frequency would increase, then skip to
-the next harmonic, then increase again, _etc._ This type of behavior
-was observed by Moore _et al._ (23) in _Aplysia_ and reported at the
-San Diego Symposium for Biomedical Electronics shortly after it was
-observed by the author in the electronic model.
-
-Thus, we have shown that an electronic analog with all parameters
-except membrane capacitance fixed at values close to those of Hodgkin
-and Huxley, can provide all of the normal threshold or axonal
-behavior and also all of the subthreshold somatic and dendritic
-behavior outlined on page 7. Whether or not this is of physiological
-significance, it certainly provides a unifying basis for construction
-of electronic neural analogs. Simple circuits, based on the
-Hodgkin-Huxley model and providing all of the aforementioned behavior,
-have been constructed with ten or fewer inexpensive transistors with
-a normal complement of associated circuitry (18). In the near future
-we hope to utilize several models of this type to help assess the
-information-processing capabilities not only of individual neurons but
-also of small groups or networks of neurons.
-
-
-REFERENCES
-
- 1. Hagiwara, S., and Bullock, T. H.
- “Intracellular Potentials in Pacemaker and Integrative Neurons of
- the Lobster Cardiac Ganglion,”
- _J. Cell and Comp. Physiol._ =50 (No. 1)=:25-48 (1957)
-
- 2. Chalazonitis, N., and Arvanitaki, A.,
- “Slow Changes during and following Repetitive Synaptic Activation
- in Ganglion Nerve Cells,”
- _Bull. Inst. Oceanogr. Monaco_ =No. 1225=:1-23 (1961)
-
- 3. Hodgkin, A. L., Huxley, A. F., and Katz, B.,
- “Measurement of Current-Voltage Relations in the Membrane of the
- Giant Axon of _Loligo_,”
- _J. Physiol._ =116=:424-448 (1952)
-
- 4. Hagiwara, S., and Saito, N.,
- “Voltage-Current Relations in Nerve Cell Membrane of Onchidium
- _verruculatum_,”
- _J. Physiol._ =148=:161-179 (1959)
-
- 5. Hagiwara, S., and Saito, N.,
- “Membrane Potential Change and Membrane Current in Supramedullary
- Nerve Cell of Puffer,”
- _J. Neurophysiol._ =22=:204-221 (1959)
-
- 6. Hagiwara, S.,
- “Current-Voltage Relations of Nerve Cell Membrane,”
- “Electrical Activity of Single Cells,”
- Igakushoin, Hongo, Tokyo (1960)
-
- 7. Bullock, T. H.,
- “Parameters of Integrative Action of the Nervous System at the
- Neuronal Level,”
- _Experimental Cell Research Suppl._ =5=:323-337 (1958)
-
- 8. Otani, T., and Bullock, T. H.,
- “Effects of Presetting the Membrane Potential of the Soma of
- Spontaneous and Integrating Ganglion Cells,”
- _Physiological Zoology_ =32 (No. 2)=:104-114 (1959)
-
- 9. Bullock, T. H., and Terzuolo, C. A.,
- “Diverse Forms of Activity in the Somata of Spontaneous and
- Integrating Ganglion Cells,”
- _J. Physiol._ =138=:343-364 (1957)
-
- 10. Bullock, T. H.,
- “Neuron Doctrine and Electrophysiology,”
- _Science_ =129 (No. 3355)=:997-1002 (1959)
-
- 11. Chalazonitis, N., and Arvanitaki, A.,
- “Slow Waves and Associated Spiking in Nerve Cells of
- _Aplysia_,”
- _Bull. Inst. Oceanogr. Monaco_ =No. 1224=:1-15 (1961)
-
- 12. Bullock, T. H.,
- “Properties of a Single Synapse in the Stellate Ganglion of
- Squid,”
- _J. Neurophysiol._ =11=:343-364 (1948)
-
- 13. Bullock, T. H.,
- “Neuronal Integrative Mechanisms,”
- “Recent Advances in Invertebrate Physiology,”
- Scheer, B. T., ed., Eugene, Oregon:Univ. Oregon Press 1957
-
- 14. Hodgkin, A. L., and Huxley, A. F.,
- “Currents Carried by Sodium and Potassium Ions through the
- Membrane of the Giant Axon of Loligo,”
- _J. Physiol._ =116=:449-472 (1952)
-
- 15. Hodgkin, A. L., and Huxley, A. F.,
- “The Components of Membrane Conductance in the Giant Axon of
- _Loligo_,”
- _J. Physiol._ =116=:473-496 (1952)
-
- 16. Hodgkin, A. L., and Huxley, A. F.,
- “The Dual Effect of Membrane Potential on Sodium Conductance in
- the Giant Axon of _Loligo_,”
- _J. Physiol._ =116=:497-506 (1952)
-
- 17. Hodgkin, A. L., and Huxley, A. F.,
- “A Quantitative Description of Membrane Current and its
- Application to Conduction and Excitation in Nerve,”
- _J. Physiol._ =117=:500-544 (1952)
-
- 18. Lewis, E. R.,
- “An Electronic Analog of the Neuron Based on the Dynamics of
- Potassium and Sodium Ion Fluxes,”
- “Neural Theory and Modeling,”
- R. F. Reiss, ed., Palo Alto, California:Stanford University
- Press, 1964
-
- 19. Eccles, J. C.,
- _Physiology of Synapses_,
- Berlin:Springer-Verlag, 1963
-
- 20. Grundfest, H.,
- “Excitation Triggers in Post-Junctional Cells,”
- “Physiological Triggers,”
- T. H. Bullock, ed., Washington, D.C.:American Physiological
- Society, 1955
-
- 21. Rall, W.,
- “Membrane Potential Transients and Membrane Time Constants of
- Motoneurons,”
- _Exp. Neurol._ =2=:503-532 (1960)
-
- 22. Araki, T., and Otani, T.,
- “The Response of Single Motoneurones to Direct Stimulation,”
- _J. Neurophysiol._ =18=:472-485 (1955)
-
- 23. Moore, G. P., Perkel, D. H., and Segundo, J. P.,
- “Stability Patterns in Interneuronal Pacemaker Regulation,”
- _Proceedings of the San Diego Symposium for Biomedical
- Engineering_, San Diego, California, 1963
-
- 24. Eccles, J. C.,
- _The Neurophysiological Basis of Mind_,
- Oxford:Clarendon Press, 1952
-
-
-
-
-Fields and Waves in Excitable Cellular Structures
-
-
- R. M. STEWART
-
- _Space General Corporation
- El Monte, California_
-
- “Study of living processes by the physiological
- method only proceeded laboriously behind the study of
- non-living systems. Knowledge about respiration, for
- instance, began to become well organized as the study
- of combustion proceeded, since this is an analogous
- operation....”
-
- J. Z. Young (24)
-
-
-INTRODUCTION
-
-The study of electrical fields in densely-packed cellular media is
-prompted primarily by a desire to understand more fully the details
-of brain mechanism and its relation to behavior. Our work has
-specifically been directed toward an attempt to model such structures
-and mechanisms, using relatively simple inorganic materials.
-
-The prototype for such experiments is the “Lillie[1] iron-wire nerve
-model.” Over a hundred years ago, it had been observed that visible
-waves were produced on the surface of a piece of iron submerged in
-nitric acid when and where the iron is touched by a piece of zinc.
-After a short period of apparent fatigue, the wire recovers and can
-again support a wave when stimulated. Major support for the idea that
-such impulses are in fact directly related to peripheral nerve impulses
-came from Lillie around 1920. Along an entirely different line,
-various persons have noted the morphological and dynamic similarity of
-dendrites in brain and those which sometimes grow by electrodeposition
-of metals from solution. Gordon Pask (17), especially, has pointed to
-this similarity and has discussed in a general way the concomitant
-possibility of a physical model for the persistent memory trace.
-
-[1] For review articles see: Lillie (13), Franck (6).
-
-By combining and extending such concepts and techniques, we hope to
-produce a macroscopic model of “gray matter,” the structural matrix of
-which will consist of a dense, homogeneously-mixed, conglomerate of
-small pellets, capable of supporting internal waves of excitation, of
-changing electrical behavior through internal fine-structure growth,
-and of forming temporal associations in response to peripheral shocks.
-
-A few experimenters have subsequently pursued the iron-wire
-nerve-impulse analogy further, hoping thereby to illuminate the
-mechanisms of nerve excitation, impulse transmission and recovery,
-but interest has generally been quite low. It has remained fairly
-undisturbed in the text books and lecture demonstrations of medical
-students, as a picturesque aid to their formal education. On the
-outer fringes of biology, still less interest has been displayed;
-the philosophical vitalists would surely be revolted by the idea of
-such models of mind and memory, and at the other end of the scale,
-contemporary computer engineers generally assume that a nerve cell
-operates much too slowly to be of any value. This lack of interest
-is certainly due, in part, to success in developing techniques of
-monitoring individual nerve fibers directly to the point that it is
-just about as easy to work with large nerve fibers (and even peripheral
-and spinal junctions) as it is to work with iron wires. Under such
-circumstances, the model has only limited value, perhaps just to the
-extent that it emphasizes the role of factors other than specific
-molecular structure and local chemical reactions in the dynamics of
-nerve action.
-
-When we leave the questions of impulse transmission on long fibers
-and peripheral junctions, however, and attempt to discuss the brain,
-there can be hardly any doubt that the development of a meaningful
-physical model technique would be of great value. Brain tissue is
-soft and sensitive, the cellular structures are small, tangled, and
-incredibly numerous. Therefore (Young (24)), “ ... physiologists hope
-that after having learned a lot about nerve-impulses in the nerves they
-will be able to go on to study how these impulses interact when they
-reach the brain. [But], we must not assume that we shall understand
-the brain only in the terms we have learned to use for the nerves.
-The function of nerves is to carry impulses—like telegraph wires. The
-functions of brains is something else.” But, confronted with such
-awesome experimental difficulties, with no comprehensive mathematical
-theory in sight, we are largely limited otherwise to verbal discourses,
-rationales and theorizing, a hopelessly clumsy tool for the development
-of an adequate understanding of brain function. A little over ten years
-ago Sperry (19) said, “Present day science is quite at a loss even
-to begin to describe the neural events involved in the simplest form
-of mental activity.” This situation has not changed much today. The
-development, study, and understanding of complex high-density cellular
-structures which incorporate characteristics of both the Lillie and
-Pask models may, it is hoped, alleviate this situation. There would
-also be fairly obvious technological applications for such techniques
-if highly developed and which, more than any other consideration, has
-prompted support for this work.
-
-Experiments to date have been devised which demonstrate the following
-basic physical functional characteristics:
-
- (1) Control of bulk resistivity of electrolytes containing
- closely-packed, poorly-conducting pellets
- (2) Circulation of regenerative waves on closed loops
- (3) Strong coupling between isolated excitable sites
- (4) Logically-complete wave interactions, including facilitation
- and annihilation
- (5) Dendrite growth by electrodeposition in “closed” excitable
- systems
- (6) Subthreshold distributed field effects, especially in
- locally-refractory regions.
-
-In addition, our attention has necessarily been directed to various
-problems of general experimental technique and choice of materials,
-especially as related to stability, fast recovery and long life.
-However, in order to understand the possible significance of, and
-motivation for such experiments, some related modern concepts of
-neurophysiology, histology and psychology will be reviewed very
-briefly. These concepts are, respectively:
-
- (1) Cellular structure in the central nervous system
- (2) Short-term or “ephemeral” memory
- (3) The synapse
- (4) Inhibition
- (5) Long-term memory traces or engram
- (6) Spatially-diffuse temporal association and learning.
-
-
-SOME CONTEMPORARY CONCEPTS
-
-Since we are attempting to duplicate processes other than chemical,
-per se, we will forego any reference to the extensive literature of
-neurochemistry. It should not be surprising though if, at the neglect
-of the fundamental biological processes of growth, reproduction and
-metabolism, it proves possible to imitate some learning mechanisms
-with grossly less complex molecular structures. There is also
-much talk of chemical versus electrical theories and mechanisms in
-neurophysiology. The distinction, when it can be made, seems to hinge
-on the question of the scale of size of significant interactions. Thus,
-“chemical” interactions presumably take place at molecular distances,
-possibly as a result of or subsequent to a certain amount of thermal
-diffusion. “Electrical” interactions, on the other hand, are generally
-understood to imply longer range or larger scale macroscopic fields.
-
-
-1. Cellular Structure
-
-The human brain contains approximately 10¹⁰ neurons to which the
-neuron theory assigns the primary role in central nervous activity.
-These cells occupy, however, a relatively small fraction of the total
-volume. There are, for example, approximately 10 times that number of
-neuroglia, cells of relatively indeterminate function. Each neuron
-(consisting of cell body, dendrites and, sometimes, an axon) comes into
-close contact with the dendrites of other neurones at some thousands
-of places, these synapses and “ephapses” being spaced approximately 5μ
-apart (1). The total number of such apparent junctions is therefore
-of the order of 10¹³. In spite of infinite fine-structure variations
-when viewed with slightly blurred vision, the cellular structure of
-the brain is remarkably homogeneous. In the cortex, at least, the
-extensions of most cells are relatively short, and when the cortex is
-at rest, it appears from the large EEG alpha-rhythms that large numbers
-of cells beat together in unison. Quoting again from Sperry, “In short,
-current brain theory encourages us to try to correlate our subjective
-psychic experience with the activity of relatively homogeneous nerve
-cell units conducting essentially homogeneous impulses, through roughly
-homogeneous cerebral tissue.”
-
-
-2. Short-Term Memory
-
-A train of impulses simply travelling on a long fiber may, for
-example, be regarded as a short-term memory much in the same way as
-a delay line acts as a transient memory in a computer. A similar
-but slightly longer term memory may also be thought of to exist in
-the form of waves circulating in closed loops (23). In fact, it is
-almost universally held today that most significant memory occurs
-in two basic interrelated ways. First of all, such a short-term
-circulating, reverberatory or regenerative memory which, however, could
-not conceivably persist through such things as coma, anesthesia,
-concussion, extreme cold, deep sleep and convulsive seizures and
-thus, secondly, a long-term memory trace which must somehow reside
-in a semipermanent fine-structural change. As Hebb (9) stated, “A
-reverbratory trace might cooperate with a structural change and carry
-the memory until the growth change is made.”
-
-
-3. The Synapse
-
-The current most highly regarded specific conception of the synapse
-is largely due to and has been best described by Eccles (5): “ ...
-the synaptic connections between nerve cells are the only functional
-connections of any significance. These synapses are of two types,
-excitatory and inhibitory, the former type tending to make nerve cells
-discharge impulses, the other to suppress the discharge. There is now
-convincing evidence that in vertebrate synapses each type operates
-through specific chemical transmitter substances ...”. In response to
-a presentation by Hebb (10), Eccles was quoted as saying, “One final
-point, and that is if there is electrical interaction, and we have seen
-from Dr. Estable’s work the complexity of connections, and we now know
-from the electronmicroscopists that there is no free space, only 200
-Å clefts, everywhere in the central nervous system, then everything
-should be electrically interacted with everything else. I think this is
-only electrical background noise and, that when we lift with specific
-chemical connections above that noise we get a significant operational
-system. I would say that there is electrical interaction but it is just
-a noise, a nuisance.” Eccles’ conclusions are primarily based on data
-obtained in the peripheral nervous system and the spinal cord. But
-there is overwhelming reason to expect that cellular interactions in
-the brain are an entirely different affair. For example, “The highest
-centres in the octopus, as in vertebrates and arthropods, contain many
-small neurons. This finding is such a commonplace, that we have perhaps
-failed in the past to make the fullest inquiry into its implications.
-Many of these small cells possess numerous processes, but no axon. It
-is difficult to see, therefore, that their function can be conductive
-in the ordinary sense. Most of our ideas about nervous functioning are
-based on the assumption that each neuron acts essentially as a link in
-some chain of conduction, but there is really no warrant for this in
-the case of cells with many short branches. Until we know more of the
-relations of these processes to each other in the neuropile it would
-be unwise to say more. It is possible that the effective part of the
-discharge of such cells is not as it is in conduction in long pathways,
-the internal circuit that returns through the same fiber, but the
-external circuit that enters other processes, ...” (3).
-
-
-4. Inhibition
-
-The inhibitory chemical transmitter substance postulated by Eccles
-has never been detected in spite of numerous efforts to do so. The
-mechanism(s) of inhibition is perhaps the key to the question of
-cellular interaction and, in one form or another, must be accounted for
-in any adequate theory.
-
-Other rather specific forms of excitation and inhibition interaction
-have been proposed at one time or another. Perhaps the best example is
-the polar neuron of Gesell (8) and, more recently, Retzlaff (18). In
-such a concept, excitatory and inhibitory couplings differ basically
-because of a macroscopic structural difference at the cellular level;
-that is, various arrangements or orientation of intimate cellular
-structures give rise to either excitation or inhibition.
-
-
-5. Long-Term Memory
-
-Most modern theories of semipermanent structural change (or _engrams_,
-as they are sometimes called) look either to the molecular level or to
-the cellular level. Various specific locales for the engram have been
-suggested, including (1) modifications of RNA molecular structure,
-(2) changes of cell size, synapse area or dendrite extensions, (3)
-neuropile modification, and (4) local changes in the cell membrane.
-There is, in fact, rather direct evidence of the growth of neurons or
-their dendrites with use and the diminution or atrophy of dendrites
-with disuse. The apical dendrite of pyramidal neurones becomes thicker
-and more twisted with continuing activity, nerve fibers swell when
-active, sprout additional branches (at least in the spinal cord) and
-presumably increase the size and number of their terminal knobs.
-As pointed out by Konorski (11), the morphological conception of
-plasticity according to which plastic changes would be related to the
-formation and multiplication of new synaptic junctions goes back at
-least as far as Ramon y Cajal in 1904. Whatever the substrate of the
-memory trace, it is, at least in adults, remarkably immune to extensive
-brain damage and as Young (24) has said: “ ... this question of the
-nature of the memory trace is one of the most obscure and disputed in
-the whole of biology.”
-
-
-6. Field Effects and Learning
-
-First, from Boycott and Young (3), “The current conception, on which
-most discussions of learning still concentrate, is that the nervous
-system consists essentially of an aggregate of chains of conductors,
-linked at key points by synapses. This reflex conception, springing
-probably from Cartesian theory and method, has no doubt proved of
-outstanding value in helping us to analyse the actions of the spinal
-cord, but it can be argued that it has actually obstructed the
-development of understanding of cerebral function.”
-
-Most observable evidence of learning and memory is extremely complex
-and its interpretation full of traps. Learning in its broadest sense
-might be detected as a semipermanent change of behavior pattern brought
-about as a result of experience. Within that kind of definition, we
-can surely identify several distinctly different types of learning,
-presumably with distinctly different kinds of mechanisms associated
-with each one. But, if we are to stick by our definition of a condition
-of semipermanent change of behavior as a criterion for learning, then
-we may also be misled into considering the development of a neurosis,
-for example, as learning, or even a deep coma as learning.
-
-When we come to consider field effects, current theories tend to get
-fairly obscure, but there seems to be an almost universal recognition
-of the fact that such fields are significant. For example, Morrell
-(16) says in his review of electrophysiological contributions to the
-neural basis of learning, “A growing body of knowledge (see reviews
-by Purpura, Grundfest, and Bishop) suggests that the most significant
-integrative work of the central nervous system is carried on in graded
-response elements—elements in which the degree of reaction depends upon
-stimulus intensity and is not all-or-none, which have no refractory
-period and in which continuously varying potential changes of either
-sign occur and mix and algebraically sum.” Gerard (7) also makes a
-number of general comments along these lines. “These attributes of
-a given cell are, in turn, normally controlled by impulses arising
-from other regions, by fields surrounding them—both electric and
-chemical—electric and chemical fields can strongly influence the
-interaction of neurones. This has been amply expounded in the case of
-the electric fields.”
-
-Learning situations involving “punishment” and “reward” or,
-subjectively, “pain” and “pleasure” may very likely be associated
-with transient but structurally widespread field effects. States of
-distress and of success seem to exert a lasting influence on behavior
-only in relation to _simultaneous_ sensory events or, better yet,
-sensory events just immediately _preceding_ in time. For example, the
-“anticipatory” nature of a conditioned reflex has been widely noted
-(21). From a structural point of view, it is as if recently active
-sites regardless of location or function were especially sensitive to
-extensive fields. There is a known inherent electrical property of both
-nerve membrane and passive iron surface that could hold the answer to
-this mechanism of spatially-diffuse temporal association; namely, the
-surface resistance drops to less than 1 per cent of its resting value
-during the refractory period which immediately follows activation.
-
-
-EXPERIMENTAL TECHNIQUE
-
-In almost all experiments, the basic signal-energy mechanism employed
-has been essentially that one studied most extensively by Lillie (12),
-Bonhoeffer (2), Yamagiwa (22), Matumoto and Goto (14) and others,
-_i.e._, activation, impulse propagation and recovery on the normally
-passive surface of a piece of iron immersed in nitric acid or of
-cobalt in chromic acid (20). The iron we have used most frequently
-is of about 99.99% purity, which gives performance more consistent
-than but similar to that obtained using cleaned “coat-hanger” wires.
-The acid used most frequently by us is about 53-55% aqueous solution
-by weight, substantially more dilute than that predominantly used by
-previous investigators. The most frequently reported concentration has
-been 68-70%, a solution which is quite stable and, hence, much easier
-to work with in open containers than the weaker solutions, results in
-very fast waves but gives, at room temperatures, a very long refractory
-period (typically, 15 minutes). A noble metal (such as silver, gold
-or platinum) placed in contact with the surface of the iron has a
-stabilizing effect (14) presumably through the action of local currents
-and provides a simple and useful technique whereby, with dilution,
-both stability and fast recovery (1 second) can be achieved in simple
-demonstrations and experiments.
-
-Experiments involving the growth by electrodeposition and study of
-metallic dendrites are done with an eye toward electrical, physical
-and chemical compatibility with the energy-producing system outlined
-above. Best results to date (from the standpoints of stability,
-non-reactivity, and morphological similarity to neurological
-structures) have been obtained by dissolving various amounts of gold
-chloride salt in 53-55% HNO₃.
-
-An apparatus has been devised and assembled for the purpose of
-containing and controlling our primary experiments. (See Figure 1).
-Its two major components are a test chamber (on the left in Figure
-1) and a fluid exchanger (on the right). In normal operation the
-test chamber, which is very rigid and well sealed after placing the
-experimental assembly inside, is completely filled with electrolyte
-(or, initially, an inert fluid) to the exclusion of all air pockets and
-bubbles. Thus encapsulated, it is possible to perform experiments which
-would otherwise be impossible due to instability. The instability which
-plagues such experiments is manifested in copious generation of bubbles
-on and subsequent rapid disintegration of all “excitable” material
-(_i.e._, iron). Preliminary experiments indicated that such “bubble
-instability” could be suppressed by constraining the volume available
-to expansion. In particular, response and recovery times can now be
-decreased substantially and work can proceed with complex systems of
-interest such as aggregates containing many small iron pellets.
-
-The test chamber is provided with a heater (and thermostatic control)
-which makes possible electrochemical impulse response and recovery
-times comparable to those of the nervous system (1 to 10 msec). The
-fluid-exchanger is so arranged that fluid in the test chamber can be
-arbitrarily changed or renewed by exchange within a rigid, sealed,
-completely liquid-filled (“isochoric”) loop. Thus, stability can
-be maintained for long periods of time and over a wide variety of
-investigative or operating conditions.
-
-Most of the parts of this apparatus are made of stainless steel and
-are sealed with polyethylene and teflon. There is a small quartz
-observation window on the test chamber, two small lighting ports, a
-pressure transducer, thermocouple, screw-and-piston pressure actuator
-and umbilical connector for experimental electrical inputs and outputs.
-
-
-BASIC EXPERIMENTS
-
-The basic types of experiments described in the following sections
-are numbered for comparison to correspond roughly to related
-neurophysiological concepts summarized in the previous section.
-
-
-1. Cellular Structure
-
-The primary object of our research is the control and determination of
-dynamic behavior in response to electrical stimulation in close-packed
-aggregates of small pellets submerged in electrolyte. Typically, the
-aggregate contains (among other things) iron and the electrolyte
-contains nitric acid, this combination making possible the propagation
-of electrochemical surface waves of excitation through the body of
-the aggregate similar to those of the Lillie iron-wire nerve model.
-The iron pellets are imbedded in and supported by a matrix of small
-dielectric (such as glass) pellets. Furthermore, with the addition
-of soluble salts of various noble metals to the electrolyte, long
-interstitial dendritic or fibrous structures of the second metal can
-be formed whose length and distribution change by electrodeposition in
-response to either internal or externally generated fields.
-
-[Illustration: Figure 1—Test chamber and fluid exchanger]
-
-Coupling between isolated excitable (iron) sites is greatly affected
-by the fine structure and effective bulk resistivity of the glass and
-fluid medium which supports and fills the space between such sites.
-In general (see Section 3, following) it is necessary, to promote
-strong coupling between small structures, to impede the “short-circuit”
-return flow of current from an active or excited surface, through
-the electrolyte and back through the dendritic structure attached
-to the same excitable site. This calls for control (increase) of
-the bulk resistivity, preferably by means specifically independent
-of electrolyte composition, which relates to and affects surface
-phenomena such as recovery (_i.e._, the “refractory” period). Figure 2
-illustrates the way in which this is being done, _i.e._, by appropriate
-choice of particle size distributions. The case illustrated shows
-the approximate proper volume ratios for maximum resistivity in a
-two-size-phase random mixture of spheres.
-
-
-2. Regenerative Loops
-
-Figure 3 shows an iron loop (about 2-inch diameter) wrapped with a
-silver wire helix which is quite stable in 53-55% acid and which
-will easily support a circulating pattern of three impulses. For
-demonstration, unilateral waves can be generated by first touching the
-iron with a piece of zinc (which produces two oppositely travelling
-waves) and then blocking one of them with a piece of platinum or a
-small platinum screen attached to the end of a stick or wand. Carbon
-blocks may also be used for this purpose.
-
-The smallest regenerative or reverberatory loop which we are at present
-able to devise is about 1 mm in diameter. Multiple waves, as expected,
-produce stable patterns in which all impulses are equally spaced. This
-phenomenon can be related to the slightly slower speed characteristic
-of the relative refractory period as compared with a more fully
-recovered zone.
-
-[Illustration]
-
-[Illustration: Figure 2—Conductivity control—mixed pellet-size
-aggregates]
-
-[Illustration: Figure 3—Regenerative or reverberatory loop]
-
-
-3. Strong Coupling
-
-If two touching pieces of iron are placed in a bath of nitric acid, a
-wave generated on one will ordinarily spread to the other. As is to be
-expected, a similar result is obtained if the two pieces are connected
-through an external conducting wire. However, if they are isolated,
-strong coupling does not ordinarily occur, especially if the elements
-are small in comparison with a “critical size,” σ/ρ where σ is the
-surface resistivity of passive iron surface (in Ω-cm²) and ρ is the
-volume resistivity of the acid (in Ω-cm). A simple and informative
-structure which demonstrates the essential conditions for strong
-electrical coupling between isolated elements of very small size may
-be constructed as shown in Figure 4. The dielectric barrier insures
-that charge transfer through one dipole must be accompanied by an equal
-and opposite transfer through the surfaces of the other dipole. If the
-“inexcitable” silver tails have sufficiently high conductance (_i.e._,
-sufficiently large surface area, hence preferably, dendrites), strong
-coupling will occur, just as though the cores of the two pieces of iron
-were connected with a solid conducting wire.
-
-[Illustration: Figure 4]
-
-[Illustration: Figure 5—Electrochemical excitatory-inhibitory
-interaction cell]
-
-
-4. Inhibitory Coupling
-
-If a third “dipole” is inserted through the dielectric membrane in
-the opposite direction, then excitation of this isolated element
-tends to inhibit the response which would otherwise be elicited by
-excitation of one of the parallel dipoles. Figure 5 shows the first
-such “logically-complete” interaction cell successfully constructed and
-demonstrated. It may be said to behave as an elementary McCulloch-Pitts
-neuron (15). Further analysis shows that similar structures
-incorporating many dipoles (both excitatory and inhibitory) can be made
-to behave as general “linear decision functions” in which all input
-weights are approximately proportional to the total size or length of
-their corresponding attached dendritic structures.
-
-
-5. Dendrite Growth
-
-Figure 6 shows a sample gold dendrite grown by electrodeposition
-(actual size, about 1 mm) from a 54% nitric acid solution to which gold
-chloride was added. When such a dendrite is attached to a piece of
-iron (both submerged), activation of the excitable element produces a
-field in such a direction as to promote further growth of the dendritic
-structure. Thus, if gold chloride is added to the solution used in
-the elementary interaction cells described above, all input influence
-“weights” tend to increase with use and, hence, produce a plasticity of
-function.
-
-
-6. Field Effects in Locally-Refractory Regions
-
-Our measurements indicate that, during the refractory period following
-excitation, the surface resistance of iron in nitric acid drops to
-substantially less than 1% of its resting value in a manner reminiscent
-of nerve membranes (4). Thus, if a distributed or gross field exists
-at any time throughout a complex cellular aggregate, concomitant
-current densities in locally-refractive regions will be substantially
-higher than elsewhere and, if conditions appropriate to dendrite
-growth exist (as described above) growth rates in such regions will
-also be substantially higher than elsewhere. It would appear that, as
-a result, recently active functional couplings (in contrast to those
-not associated with recent neural activity) should be significantly
-altered by widely distributed fields or massive peripheral shocks. This
-mechanism might thus explain the apparent ability of the brain to form
-specific temporal associations in response to spatially-diffuse effects
-such as are generated, for example, by the pain receptors.
-
-[Illustration: (a)]
-
-[Illustration: (b)
-
-Figure 6—Dendritic structures, living and non-living. (a) Cat dendrite
-trees (from Bok, “Histonomy of the Cerebral Cortex,” Elsevier, 1959);
-(b) Electrodeposited gold dendrite tree.]
-
-
-SUMMARY
-
-An attempt is being made to develop meaningful electrochemical model
-techniques which may contribute toward a clearer understanding of
-cortical function. Two basic phenomena are simultaneously employed
-which are variants of (1) the Lillie iron-wire nerve model, and (2)
-growth of metallic dendrites by electrodeposition. These phenomena are
-being induced particularly within dense cellular aggregates of various
-materials whose interstitial spaces are flooded with liquid electrolyte.
-
-
-REFERENCES
-
- 1. Bok, S. T.,
- “Histonomy of the Cerebral Cortex,”
- Amsterdam, London:Elsevier Publishing Co., New York:Princeton,
- 1959
-
- 2. Bonhoeffer, K. F.,
- “Activation of Passive Iron as a Model for the Excitation of
- Nerve,”
- _J. Gen. Physiol._ =32=:69-91 (1948).
- This paper summarizes work carried out during 1941-1946
- at the University of Leipzig, and published during the
- war years in German periodicals.
-
- 3. Boycott, B. B., and Young, J. Z.,
- “The Comparative Study of Learning,”
- S. E. B. Symposia, No. IV
- “Physiological Mechanisms in Animal Behavior,”
- Cambridge: University Press, USA:Academic Press, Inc., 1950
-
- 4. Cole, K. S., and Curtis, H. J.,
- “Electric Impedance of the Squid Giant Axon During Activity,”
- _J. Gen. Physiol._ =22=:649-670 (1939)
-
- 5. Eccles, J. C.,
- “The Effects of Use and Disuse of Synaptic Function,”
- “Brain Mechanisms and Learning—A Symposium,”
- organized by the Council for International Organizations of
- Medical Science, Oxford:Blackwell Scientific Publications, 1961
-
- 6. Franck, U. F.,
- “Models for Biological Excitation Processes,”
- “Progress in Biophysics and Biophysical Chemistry,”
- J. A. V. Butler, ed., London and New York:Pergamon Press,
- pp. 171-206, 1956
-
- 7. Gerard, R. W.,
- “Biological Roots of Psychiatry,”
- _Science_ =122 (No. 3162)=:225-230 (1955)
-
- 8. Gesell, R.,
- “A Neurophysiological Interpretation of the Respiratory Act,”
- _Ergedn. Physiol._ =43:=477-639 (1940)
-
- 9. Hebb, D. O.,
- “The Organization of Behavior, A Neuropsychological Theory,”
- New York:John Wiley and Sons, 1949
-
- 10. Hebb, D. O.,
- “Distinctive Features of Learning in the Higher Animal,”
- “Brain Mechanisms and Learning—A Symposium,”
- organized by the Council for International Organizations of
- Medical Science, Oxford:Blackwell Scientific Publications, 1961
-
- 11. Konorski, J.,
- “Conditioned Reflexes and Neuron Organization,”
- Cambridge:Cambridge University Press, 1948
-
- 12. Lillie, R. S.,
- “Factors Affecting the Transmission and Recovery in the Passive
- Iron Nerve Model,”
- _J. Gen. Physiol._ =4=:473 (1925)
-
- 13. Lillie, R. S., _Biol. Rev._ =16=:216 (1936)
-
- 14. Matumoto, M., and Goto, K.,
- “A New Type of Nerve Conduction Model,”
- _The Gurma Journal of Medical Sciences_ =4(No. 1)= (1955)
-
- 15. McCulloch, W. S., and Pitts, W.,
- “A Logical Calculus of the Ideas Immanent in Nervous Activity,”
- _Bulletin of Mathematical Biophysics_ =5=:115-133 (1943)
-
- 16. Morrell, F.,
- “Electrophysiological Contributions to the Neural
- Basis of Learning,”
- _Physiological Reviews_ =41(No. 3)= (1961)
-
- 17. Pask, G.,
- “The Growth Process Inside the Cybernetic Machine,”
- _Proc. 2nd Congress International Association Cybernetics_,
- Gauthier-Villars, Paris:Namur, 1958
-
- 18. Retzlaff, E.,
- “Neurohistological Basis for the Functioning of Paired
- Half-Centers,”
- _J. Comp. Neurology_ =101=:407-443 (1954)
-
- 19. Sperry, R. W.,
- “Neurology and the Mind-Brain Problem,”
- _Amer. Scientist_ =40(No. 2)=: 291-312 (1952)
-
- 20. Tasaki, I., and Bak, A. F.,
- _J. Gen. Physiol._ =42=:899 (1959)
-
- 21. Thorpe, W. H.,
- “The Concepts of Learning and Their Relation to Those of
- Instinct,”
- S. E. B. Symposia, No. IV,
- “Physiological Mechanisms in Animal Behavior,”
- Cambridge:University Press, USA:Academic Press, Inc., 1950
-
- 22. Yamagiwa, K.,
- “The Interaction in Various Manifestations
- (Observations on Lillie’s Nerve Model),”
- _Jap. J. Physiol._ =1=:40-54 (1950)
-
- 23. Young, J. Z.,
- “The Evolution of the Nervous System and of the
- Relationship of Organism and Environment,”
- G. R. de Beer, ed.,
- “Evolution,”
- Oxford:Clarendon Press, pp. 179-204, 1938
-
- 24. Young, J. Z.,
- “Doubt and Certainty in Science, A Biologist’s Reflections
- on the Brain,”
- New York:Oxford Press, 1951
-
-
-
-
-Multi-Layer Learning Networks
-
-
- R. A. STAFFORD
-
- _Philco Corp., Aeronutronic Division
- Newport Beach, California_
-
-
-INTRODUCTION
-
-This paper is concerned with the problem of designing a network of
-linear threshold elements capable of efficiently adapting its various
-sets of weights so as to produce a prescribed input-output relation.
-It is to accomplish this adaptation by being repetitively presented
-with the various inputs along with the corresponding desired outputs.
-We will not be concerned here with the further requirement of various
-kinds of ability to “generalize”—_i.e._, to tend to give correct
-outputs for inputs that have not previously occurred when they are
-similar in some transformed sense to other inputs that have occurred.
-
-In putting forth a model for such an adapting or “learning” network, a
-requirement is laid down that the complexity of the adaption process
-in terms of interconnections among elements needed for producing
-appropriate weight changes, should not greatly exceed that already
-required to produce outputs from inputs with a static set of weights.
-In fact, it has been found possible to use the output-from-input
-computing capacity of the network to help choose proper weight changes
-by observing the effect on the output of a variety of possible weight
-changes.
-
-No attempt is made here to defend the proposed network model on
-theoretical grounds since no effective theory is known at present.
-Instead, the plausibility of the various aspects of the network model,
-combined with empirical results must suffice.
-
-
-SINGLE ELEMENTS
-
-To simplify the problem it is assumed that the network receives a set
-of two-valued inputs, x₁, x₂, ..., xₙ, and is required to produce only
-a single two-valued output, y. It is convenient to assign the numerical
-quantities +1 and -1 to the two values of each variable.
-
-The simplest network would consist of a single linear threshold
-element with a set of weights, c₀, c₁, c₂, ..., cₙ. These determine
-the output-input relation or function so that y is +1 or -1 according
-as the quantity, c₀ + c₁x₁ + c₂x₂ + ... + cₙxₙ, is positive or not,
-respectively. It is possible for such a single element to exhibit an
-adaptive behavior as follows. If, for a given set, x₁, x₂, ..., xₙ, the
-output, y, is correct, then make no changes to the weights. Otherwise
-change the weights according to the equations
-
-Δc₀ = y* Δcᵢ = y*xᵢ, i = 1,2, ...,n
-
-where y* is the desired output.
-
-It has been shown by a number of people that the weights of such an
-element are assured of arriving at a set of values which produce the
-correct output-input relation after a sufficient number of errors,
-provided that such a set exists. An upper bound on the number of
-possible errors can be given which depends only on the initial weight
-values and the logical function to be learned. This does not, however,
-solve our network problem for two reasons.
-
-First, as the number, n, of inputs gets large, the number of errors
-to be expected for most functions which can be learned increases to
-unreasonable values. For example, for n = 6, most such functions
-result in 500 to 1000 errors compared to an average of 32 errors to be
-expected in a perfect learning device.
-
-Second, and more important, the fraction of those logical functions
-which can be generated in a single element becomes vanishingly small as
-n increases. For example, at n = 6 less than one in each three trillion
-logical functions is so obtainable.
-
-
-NETWORKS OF ELEMENTS
-
-It can be demonstrated that if a sufficiently large number of linear
-threshold elements is used, with the outputs of some being the inputs
-of others, then a final output can be produced which is any desired
-logical function of the inputs. The difficulty in such a network lies
-in the fact that we are no longer provided with a knowledge of the
-correct output for each element, but only for the final output. If the
-final output is incorrect there is no obvious way to determine which
-sets of weights should be altered.
-
-As a result of considerable study and experimentation at Aeronutronic,
-a network model has been evolved which, it is felt, will get around
-these difficulties. It consists of four basic features which will now
-be described.
-
-
-Positive Interconnecting Weights
-
-It is proposed that all weights in elements attached to inputs which
-come from other elements in the network be restricted to positive
-values. (Weights attached to the original inputs to the network, of
-course, must be allowed to be of either sign.) The reason for such a
-restriction is this. If element 1 is an input to element 2 with weight
-c₁₂, element 2 to element 3 with weight c₂₃, _etc._, then the sign of
-the product, c₁₂c₂₃ ..., gives the sense of the effect of a change in
-the output of element 1 on the final element in the chain (assuming
-this is the only such chain between the two elements). If these various
-weights were of either possible sign, then a decision as to whether or
-not to change the output in element 1 to help correct an error in the
-final element would involve all weights in the chain. Moreover, since
-there would in general be a multiplicity of such chains, the decision
-is rendered impossibly difficult.
-
-The above restriction removes this difficulty. If the output of any
-element in the network is changed, say, from -1 to +1, the effect on
-the final element, if it is affected at all, is in the same direction.
-
-It should be noted that this restriction does not seriously affect
-the logical capabilities of a network. In fact, if a certain logical
-function can be achieved in a network with the use of weights of
-unrestricted sign, then the same function can be generated in another
-network with only positive interconnecting weights and, at worst, twice
-the number of elements. In the worst case this is done by generating
-in the restricted network both the output and its complement for each
-element of the unrestricted network. (It is assumed that there are no
-loops in the network.)
-
-
-A Variable Bias
-
-The central problem in network learning is that of determining, for
-a given input, the set of elements whose outputs can be altered so
-as to correct the final element, and which will do the least amount
-of damage to previous adaptations to other inputs. Once this set has
-been determined, the incrementing rule given for a single element will
-apply in this case as well (subject to the restriction of leaving
-interconnecting weights positive), since the desired final output
-coincides with that desired for each of the elements to be changed
-(because of positive interconnecting weights).
-
-In the process of arriving at such a decision three factors need to be
-considered. Elements selected for change should tend to be those whose
-output would thereby be affected for a minimum number of other possible
-inputs. At the same time it should be ascertained that a change in
-each of the elements in question does indeed contribute significantly
-towards correcting the final output. Finally, a minimum number of such
-elements should be used.
-
-It would appear at first that this kind of decision is impossible to
-achieve if the complexity of the decision apparatus is kept comparable
-to that of the basic input-output network as mentioned earlier.
-However, in the method to be described it is felt that a reasonable
-approximation to these requirements will be achieved without an undue
-increase in complexity.
-
-It is assumed that in addition to its normal inputs, each element
-receives a variable input bias which we can call b. The output of every
-element should then be determined by the sign of the usual weighted
-sum of its inputs plus this bias quantity. This bias is to be the same
-for each element of the network. If b = 0 the network will behave
-as before. However, if b is increased gradually, various elements
-throughout the network will commence changing from -1 to +1, with one
-or a few changing at any one time as a rule. If b is decreased, the
-opposite will occur.
-
-Now suppose that for a given input the final output ought to be +1 but
-actually is -1. Assume that b is then raised so high that this final
-output is corrected. Then commence a gradual decline in b. Various
-elements may revert to -1, but until the final output does, no weights
-are changed. When the final output does revert to -1, it is due to an
-element’s having a sum (weighted sum plus bias) which just passed down
-through zero. This then caused a chain effect of changing elements
-up to the final element, but presumably this element is the only one
-possessing a zero sum. This can then be the signal for the weights
-on an element to change—a change of final output from right to wrong
-accompanied simultaneously by a zero sum in the element itself.
-
-After such a weight change, the final output will be correct once more
-and the bias can again proceed to fall. Before it reaches zero, this
-process may occur a number of times throughout the network. When the
-bias finally stands at zero with the final output correct, the network
-is ready for the next input. Of course if -1 is desired, the bias will
-change in the opposite direction.
-
-It is possible that extending the weight change process a little past
-the zero bias level may have beneficial results. This might increase
-the life expectancy of each learned input-output combination and
-thereby reduce the total number of errors. This is because the method
-used above can stop the weight correction process so that even though
-the final output is correct, some elements whose output are essential
-to the final output have sums close to zero, which are easily changed
-by subsequent weight changes.
-
-It will be noted that this method conforms to all three considerations
-mentioned previously. First, by furnishing each element the same bias,
-and by not changing weights until the final output becomes incorrect
-with dropping bias, there is a strong tendency to select elements
-which, with b = 0, would have sums close to zero. But the size of the
-sum in an element is a good measure of the amount of damage done to
-an element for other inputs if its current output is to be changed.
-Second, it is obvious that each element changed has had a demonstrable
-effect on the final output. Finally, there will be a clear tendency to
-change only a minimum of elements because changes never occur until the
-output clearly requires a change.
-
-On the other hand this method requires little more added complexity to
-the network than it already has. Each element requires a bias, an error
-signal, and the desired final output, these things being uniform for
-all elements in a network. Some external device must manipulate the
-bias properly, but this is a simple behavior depending only on an error
-signal and the desired final output—not on the state of individual
-elements in the network. What one has, then, is a network consisting
-of elements which are nearly autonomous as regards their decisions
-to change weights. Such a scheme appears to be the only way to avoid
-constructing a central weight-change decision apparatus of great
-complexity. This rather sophisticated decision is made possible by
-utilizing the computational capabilities the network already possesses
-in producing outputs from inputs.
-
-It should be noted here that this varying bias method requires that
-the variable bias be furnished to just those elements which have
-variable weights and to no others. Any fixed portion of the network,
-such as preliminary layers or final majority function for example,
-must operate independently of the variable bias. Otherwise, the final
-output may go from right to wrong as the bias moves towards zero and no
-variable-weight element be to blame. In such a case the network would
-be hung up.
-
-
-Logical Redundancy in the Network
-
-A third aspect of the network model is that for all the care taken
-in the previous steps, they will not suffice in settling quickly to
-a set of weights that will generate the required logical function
-unless there is a great multiplicity of ways in which this can be done.
-This is to say that a learning network needs to have an excess margin
-of weights and elements beyond the minimum required to generate the
-functions which are to be learned.
-
-This is analogous to the situation that prevails for a single element
-as regards the allowed range of values on its weights. It can be shown
-for example, that any function for n=6 that can be generated by a
-single element can be obtained with each weight restricted to the range
-of integer values -9,-8, ..., +9. Yet no modification of the stated
-weight change rule is known which restricts weight values to these and
-yet has any chance of ever being learned for most functions.
-
-
-Fatigued Elements
-
-It would appear from some of the preliminary results of network
-simulations that it may be useful to have elements become “fatigued”
-after undergoing an excessive number of weight changes. Experiments
-have been performed on simplifications of the model described so far
-which had the occasional result that a small number of elements came
-to a state where they received most of the weight increments, much
-to the detriment of the learning process. In such cases the network
-behaves as if it were composed of many fewer adjustable elements. In a
-sense this is asking each element to maintain a record of the data it
-is being asked to store so that it does not attempt to exceed its own
-information capacity.
-
-It is not certain just how this fatigue factor should enter in the
-element’s actions, but if it is to be compatible with the variable bias
-method, this fatigue factor must enter into the element’s response to
-a changing bias. Once an element changes state with zero sum at the
-same time that the final output becomes wrong, incrementing must occur
-if the method is to work. Hence a “fatigued” element must respond less
-energetically to a change of bias, perhaps with a kind of variable
-factor to be multiplied by the bias term.
-
-
-NETWORK STRUCTURE
-
-It is felt that the problem of selecting the structure of
-interconnections for a network is intimately connected to the
-previously mentioned problem of generalization. Presumably a given
-type of generalization can be obtained by providing appropriate fixed
-portions of the network and an appropriate interconnection structure
-for the variable portion. However, for very large networks, it is
-undoubtedly necessary to restrict the complexity so that it can be
-specified by relatively simple rules. Since very little is known about
-this quite important problem, no further discussion will be attempted
-here.
-
-
-COMPUTER SIMULATION RESULTS
-
-A computer simulation of some of the network features previously
-described has been made on an IBM 7090. Networks with an excess of
-elements and with only positive interconnecting weights were used.
-However, in place of the variable bias method, a simple choice of the
-element of sum closest to, and on the wrong side of, zero was made
-without regard to the effectiveness of the element in correcting the
-final output. No fatigue factors were used.
-
-The results of these simulations are very encouraging, but at the same
-time indicate the need for the more sophisticated methods. No attempt
-will be made here to describe the results completely.
-
-In one series of learning experiments, a 22-element network was used
-which had three layers, 10 elements on the first, 11 on the second, and
-1 on the third. The single element on the third was the final output,
-and was a fixed majority function of the 11 elements in the second
-layer. These in turn each received inputs from each of the 10 on the
-first layer and from each of the 6 basic inputs. The 10 on the first
-layer each received only the 6 basic inputs. A set of four logical
-functions, A, B, C, and D, was used. Function A was actually a linear
-threshold function which could be generated by the weights 8, 7, 6, 5,
-4, 3, 2, functions B and C were chosen by randomly filling in a truth
-table, while D was the parity function.
-
- TABLE I
- -------+--------+---------+-------
- A | B | C | D
- r e | r e | r e | r e
- -------+--------+---------+-------
- 5 54 | 8 100 | 11 101 | 4 52
- 4 37 | 9 85 | 4 60 | 5 62
- 4 44 | 6 72 | 9 85 | 6 56
- -------+--------+---------+-------
-
-Table I gives the results of one series of runs with these functions
-and this network, starting with various random initial weights. The
-quantity, r, is the number of complete passes through the 64-entry
-truth table before the function was completely learned, while e is
-the total number of errors made. In evaluating the results it should
-be noted that an ideal learning device would make an average of 32
-errors altogether on each run. The totals recorded in these runs are
-agreeably close to this ideal. As expected, the linear threshold
-function is the easiest to learn, but it is surprising that the
-parity function was substantially easier than the two randomly chosen
-functions. Table II gives a chastening result of the same experiment
-with all interconnecting weights removed except that the final element
-is a fixed majority function of the other 21 elements. Thus there was
-adaptation on one layer only. As can be seen Table I is hardly better
-than Table II so that the value of variable interconnecting weights was
-not being fully realized. In a later experiment the number of elements
-was reduced to 12 elements and the same functions used. In this case
-the presence of extra interconnecting weights actually proved to be
-a hindrance! However a close examination of the incrementing process
-brought out the fact that the troublesome behavior was due to the
-greater chance of having only a few (often only one) elements do nearly
-all the incrementing. It is expected that the use of the additional
-refinements discussed herein will produce a considerable improvement
-in bringing out the full power of adaptation in multiple layers of a
-network.
-
- TABLE II
- -------+--------+---------+-------
- A | B | C | D
- r e | r e | r e | r e
- -------+--------+---------+-------
- 7 47 | 18 192 | 8 110 | 4 48
- 3 40 | 7 69 | 10 98 | 6 68
- 4 43 | 7 82 | 4 47 | 6 46
- -------+--------+---------+-------
-
-
-FUTURE PROBLEMS
-
-Aside from the previous question of deciding on network structure,
-there are several other questions that remain to be studied in learning
-networks.
-
-There is the question of requiring more than a single output from a
-network. If, say, two outputs are required for a given input, one
-+1 and the other -1, this runs into conflict with the incrementing
-process. Changes that aid one output may act against the other.
-Apparently the searching process depicted before with a varying bias
-must be considerably refined to find weight changes which act on
-all the outputs in the required way. This is far from an academic
-question because there will undoubtedly be numerous cases in which
-the greatest part of the input-output computation will have shared
-features for all output variables. Only at later levels do they need to
-be differentiated. Hence it is necessary to envision a single network
-producing multiple outputs rather than a separate network for each
-output variable if full efficiency is to be achieved.
-
-Another related question is that of using input variables that are
-either many-, or continuous-, valued rather than two-valued. No
-fundamental difficulties are discernible in this case, but the matter
-deserves some considerable study and experimentation.
-
-Another important question involves the use of a succession of inputs
-for producing an output. That is, it may be useful to allow time to
-enter into the network’s logical action, thus giving it a “dynamic” as
-well as “static” capability.
-
-
-
-
-Adaptive Detection of Unknown Binary Waveforms
-
-
- J. J. SPILKER, JR.
-
- _Philco Western Development Laboratories
- Palo Alto, California_
-
- This work was supported by the Philco WDL
- Independent Development Program. This paper,
- submitted after the Symposium, represents a
- more detailed presentation of some of the
- issues raised in the discussion sessions at the
- Symposium and hence, constitutes a worthwhile
- addition to the Proceedings.
-
-
-INTRODUCTION
-
-One of the most important objectives in processing a stream of
-data is to determine and detect the presence of any invariant or
-quasi-invariant “features” in that data stream. These features are
-often initially unknown and must be “learned” from the observations.
-One of the simplest features of this form is a finite length signal
-which occurs repetitively, but not necessarily periodically with time,
-and has a waveshape that remains invariant or varies only slowly with
-time.
-
-In this discussion, we assume that the data stream has been
-pre-processed, perhaps by a detector or discriminator, so as to exhibit
-this type of repetitive (but unknown) waveshape or signal structure.
-The observed signal, however, is perturbed by additive noise or other
-disturbances. It is desired to separate the quasi-invariance of the
-data from the truly random environment. The repetitive waveform may
-represent, for example, the transmission of an unknown sonar or radar,
-a pulse-position modulated noise-like waveform, or a repeated code word.
-
-The problem of concern is to estimate the signal waveshape and to
-determine the time of each signal occurrence. We limit this discussion
-to the situation where only a single repetitive waveform is present
-and the signal sample values are binary. The observed waveform is
-assumed to be received at low signal-to-noise ratio so that a single
-observation of the signal (even if one knew precisely the arrival time)
-is not sufficient to provide a good estimate of the signal waveshape.
-The occurrence time of each signal is assumed to be random.
-
-
-THE ADAPTIVE DETECTION MACHINE
-
-The purpose of this note is to describe very briefly a machine[2] which
-has been implemented to recover the noise-perturbed binary waveform.
-A simplified block diagram of the machine is shown in Figure 1. The
-experimental machine has been designed to operate on signals of 10³
-samples duration.
-
-[2] The operation of this machine is described in substantially greater
-detail in J. J. Spilker, Jr., D. D. Luby, R. D. Lawhorn, “Adaptive
-Binary Waveform Detection,” Philco Western Development Laboratories,
-Communication Sciences Department, TR #75, December 1963.
-
-Each analog input sample enters the machine at left and may either
-contain a signal sample plus noise or noise alone. In order to permit
-digital operation in the machine, the samples are quantized in a
-symmetrical three-level quantizer. The samples are then converted
-to vector form, _e.g._, the previous 10³ samples form the vector
-components. A new input vector, ⮕Y⁽ⁱ⁾, is formed at each sample instant.
-
-Define the signal sample values as s₁, s₂, ..., sₙ. The observed vector
-Y⁽ⁱ⁾ is then either (a) perfectly centered signal plus noise, (b)
-shifted signal plus noise, or (c) noise alone.
-
- { (s₁, s₂, ..., sₙ) + (n₁, n₂, ..., nₙ) (a)
- (Y⁽ⁱ⁾)ᵗ = { (0, ..., s₁, s₂, ..., sₙ₋ⱼ) + (n₁, n₂, ..., nₙ) (b)
- { (0 ... 0) + (n₁, n₂, ..., nₙ) (c)
-
-At each sample instant, two measurements are made on the input
-vector, an energy measurement ‖Y⁽ⁱ⁾‖² and a polarity coincidence
-cross-correlation with the present estimate of the signal vector stored
-in memory. If the weighted sum of the energy and cross-correlation
-measurements exceeds the present threshold value Γᵢ, the input vector
-is accepted as containing the signal (properly shifted in time), and
-the input vector is added to the memory. The adaptive memory has 2^{Q}
-levels, 2^{Q-1} positive levels, 1 zero level and 2^{Q-1}-1 negative
-levels. New contributions are made to the memory by normal vector
-addition except that saturation occurs when a component value is at the
-maximum or minimum level.
-
-The acceptance or rejection of a given input vector is based on a
-hypersphere decision boundary. The input vector is accepted if the
-weighted sum γᵢ exceeds the threshold Γᵢ
-
- γᵢ = Y⁽ⁱ⁾∙M⁽ⁱ⁾ + α‖Y⁽ⁱ⁾‖² ⩾ Γᵢ.
-
-[Illustration: Figure 1—Block diagram of the adaptive binary waveform
-detector]
-
-Geometrically, we see that the input vector is accepted if it falls on
-or outside of a hypersphere centered at ⮕C⁽ⁱ⁾ = -⮕M⁽ⁱ⁾/2α having radius
-squared
-
- Γ⁽ⁱ⁾ ‖M⁽ⁱ⁾‖²
- [r⁽ⁱ⁾]² = ——— + —————— .
- α (2α)²
-
-Both the center and radius of this hypersphere change as the machine
-adapts. The performance and optimality of hypersphere-type decision
-boundaries have been _discussed in related work_ by Glaser[3] and
-Cooper.[4]
-
-[3] F. M. Glaser, “Signal Detection by Adaptive Filters,” _IRE Trans.
-Information Theory_, pp. 87-90; April 1961.
-
-[4] P. W. Cooper, “The Hypersphere in Pattern Recognition,”
-_Information and Control_, pp. 324-346; December 1962.
-
-The threshold value, Γᵢ, is adapted so that it increases if the
-memory becomes a better replica of the signal with the result that γᵢ
-increases. On the other hand, if the memory is a poor replica of the
-signal (for example, if it contains noise alone), it is necessary that
-the threshold decay with time to the point where additional acceptances
-can modify the memory structure.
-
-The experimental machine is entirely digital in operation and, as
-stated above, is capable of recovering waveforms of up to 10³ samples
-in duration. In a typical experiment, one might attempt to recover
-an unknown noise-perturbed, pseudo-random waveform of up to 10³ bits
-duration which occurs at random intervals. If no information is
-available as to the signal waveshape, the adaptive memory is blank at
-the start of the experiment.
-
-In order to illustrate the operation of the machine most clearly, let
-us consider a repetitive binary waveform which is composed of 10³ bits
-of alternate “zeros” and “ones.” A portion of this waveform is shown in
-Figure 2a. The waveform actually observed is a noise-perturbed version
-of this waveform shown in Figure 2b at-6 db signal-to-noise ratio. The
-exact sign of each of the signal bits obviously could not be accurately
-determined by direct observation of Figure 2b.
-
-[Illustration: (a) Binary signal]
-
-[Illustration: (b) Binary signal plus noise
-
-Figure 2—Binary signal with additive noise at-6 db SNR]
-
-[Illustration: (a) (b)]
-
-[Illustration: (c) (d)]
-
-[Illustration: (e)
-
-Figure 3—Adaption of the memory at-6 db SNR: (a) Blank initial memory;
-(b) Memory after first dump; (c) Memory after 12 dumps; (d) Memory
-after 40 dumps; (e) Perfect “checkerboard” memory for comparison]
-
-As the machine memory adapts to this noisy input signal, it progresses
-as shown in Figure 3. The sign of 10^{3} memory components are
-displayed in a raster pattern in this figure. Figure 3a shows the
-memory in its blank initial state at the start of the adaption process.
-Figure 3b shows the memory after the first adaption of the memory. This
-first “dump” occurred after the threshold had decayed to the point
-where an energy measurement produced an acceptance decision. Figure
-3c and 3d show the memory after 12 and 40 adaptions, respectively.
-These dumps, of course, are based on both energy and cross-correlation
-measurements. As can be seen, the adapted memory after 40 dumps is
-already quite close to the perfect memory shown by the “checkerboard”
-pattern of Figure 3c.
-
-The detailed analysis of the performance of this type of machine
-vs. signal-to-noise ratio, average signal repetition rate, signal
-duration, and machine parameters is extremely complex. Therefore, it
-is not appropriate here to detail the results of the analytical and
-experimental work on the performance of this machine. However, several
-conclusions of a general nature can be stated.
-
- (a) Because the machine memory is always adapting, there
- is a relatively high penalty for “false alarms.”
- False alarms can destroy a perfect memory. Hence,
- the threshold level needs to be set appropriately
- high for the memory adaption. If one wishes to
- detect signal occurrences with more tolerance to
- false alarms, a separate comparator and threshold
- level should be used.
-
- (b) The present machine structure, which allows for
- slowly varying changes in the signal waveshape,
- exhibits a marked threshold effect in steady-state
- performance at an input signal-to-noise ratio
- (peak signal power-to-average noise power ratio)
- of about -12 db. Below this signal level, the time
- required for convergence increases very rapidly with
- decreasing signal level. At higher SNR, convergence
- to noise-like signals, having good auto-correlation
- properties, occurs at a satisfactory rate.
-
-A more detailed discussion of performance has been published in the
-report cited in footnote reference 1.
-
-
-
-
-Conceptual Design of Self-Organizing Machines
-
-
- P. A. KLEYN
-
- _Northrop Nortronics_
- _Systems Support Department_
- _Anaheim, California_
-
- Self-organization is defined and several examples
- which motivate this definition are presented. The
- significance of this definition is explored by
- comparison with the metrization problem discussed
- in the companion paper (1) and it is seen that
- self-organization requires decomposing the space
- representing the environment. In the absence
- of a priori knowledge of the environment, the
- self-organizing machine must resort to a sequence
- of projections on unit spheres to effect this
- decomposition. Such a sequence of projections
- can be provided by repeated use of a nilpotent
- projection operator (NPO). An analog computer
- mechanization of one such NPO is discussed
- and the signal processing behavior of the NPO
- is presented in detail using the Euclidean
- geometrical representation of the metrizable
- topology provided in the companion paper.
- Self-organizing systems using multiple NPO’s
- are discussed and current areas of research are
- identified.
-
-
-INTRODUCTION
-
-Unlike the companion paper which considers certain questions in
-depth, this paper presents a survey of the scope of our work in
-self-organizing systems and is not intended to be profound.
-
-The approach we have followed may be called phenomenological (Figure
-1). That is, the desired behavior (self-organization) was defined,
-represented mathematically, and a mechanism(s) required to yield the
-postulated behavior was synthesized using mathematical techniques. One
-advantage of this approach is that it avoids assumptions of uniqueness
-of the mechanism. Another advantage is that the desired behavior, which
-is after all the principal objective, is taken as invariant. An obvious
-disadvantage is the requirement for the aforementioned synthesis
-technique; fortunately in our case a sufficiently general technique had
-been developed by the author of the companion paper.
-
-From the foregoing and from the definition of self-organization we
-employ (see conceptual model), it would appear that our research does
-not fit comfortably within any of the well publicized approaches to
-self-organization (2). Philosophically, we lean toward viewpoints
-expressed by Ashby (3), (4), Hawkins (5), and Mesarovic (6) but with
-certain reservations. We have avoided the neural net approach partly
-because it is receiving considerable attention and also because the
-brain mechanism need not be the unique way to produce the desired
-behavior.
-
-[Illustration: Figure 1—Approach used in Nortronics research on
-self-organizing systems]
-
-Nor have we followed the probability computer or statistical decision
-theory approach exemplified by Braverman (7) because these usually
-require some sort of preassigned coordinate system (8). Neither will
-the reader find much indication of formal logic (9) or heuristic (10)
-programming. Instead, we view a self-organizing system more as a mirror
-whose appearance reflects the environment rather than its own intrinsic
-nature. With this viewpoint, a self-organizing system appears very
-flexible because it possesses few internal constraints which would tend
-to distort the reflection of the environment and hinder its ability to
-adapt.
-
-
-CONCEPTUAL MODEL
-
-Definition
-
-A system is said to be self-organizing if, after observing the input
-and output of an unknown phenomenon (transfer relation), the system
-organizes itself into a simulation of the unknown phenomenon.
-
-Implicit in this definition is the requirement that the self-organizing
-machine (SOM) not possess a preassigned coordinate system. In fact it
-is just this ability to acquire that coordinate system implicit in the
-input-output spaces which define the phenomenon that we designate as
-self-organization. Thus any a priori information programmed into the
-SOM by means of, for example, stored or wired programs, constrains
-the SOM and limits its ability to adapt. We do not mean to suggest
-that such preprogramming is not useful or desirable; merely that it is
-inconsistent with the requirement for self-organization. As shown in
-Figure 2, it is the given portion of the environment which the SOM is
-to simulate, which via the defining end spaces, furnishes the SOM with
-all the data it needs to construct the coordinate system intrinsic to
-those spaces.
-
-The motivation for requiring the ability to simulate as a feature of
-self-organization stems from the following examples.
-
-Consider the operation of driving an automobile. Figure 3 depicts the
-relation characterized by a set of inputs; steering, throttle, brakes,
-transmission, and a set of outputs; the trajectory. Operation of the
-automobile requires a device (SOM) which for a desired trajectory can
-furnish those inputs which realize the desired trajectory. In order to
-provide the proper inputs to the automobile, the SOM must contain a
-simulation of ⨍⁻¹(x).
-
-[Illustration: Figure 2—Simulation of (a portion of) the environment]
-
-[Illustration: Figure 3—Simulation of a relation]
-
-Since ⨍(x) is completely defined in terms of the inputs and the
-resulting trajectories, exposure to them provide the SOM with all the
-information necessary to simulate ⨍⁻¹(x). And if the SOM possesses
-internal processes which cause rearrangement of the input-output
-relation of the SOM to correspond to ⨍⁻¹(x) in accordance with the
-observed data, the SOM can operate an automobile. It is this internal
-change which is implied by the term “self-organizing,” but note that
-the instructions which specify the desired organization have their
-source in the environment.
-
-As a second example consider adaptation to the environment. Adapt
-(from Webster) means: “to change (oneself) so that one’s behavior,
-attitudes, _etc._, will conform to new or changed circumstances.
-Adaptation in biology means a change in structure, function or form
-that produces better adjustment to the environment.” These statements
-suggest a simulation because adjustment to the environment implies
-survival by exposing the organism to the beneficial rather than the
-inimical effects of the environment. If we represent the environment
-(or portion thereof) as a relation as shown in Figure 2, we note that
-the ability to predict what effect a given disturbance will have is due
-to a simulation of the cause-effect relation which characterizes the
-environment.
-
-It would be a mistake to infer from these examples that simulation
-preserves the appearance of the causes and effects which characterize
-a relation. We clarify this situation by examining a relation and its
-simulation.
-
-Consider the relation between two mothers and their sons as pictured
-in Figure 4. Observe that if symbols (points) are substituted for the
-actual physical objects (mothers and sons), the relation is not altered
-in any way. This is what we mean by simulation and this is how a SOM
-simulates. It is not even necessary that the objects, used to display
-the relation, be defined; _i.e._, these objects may be primitive.
-(If this were not so, no mathematical or physical theory could model
-the environment.) The main prerequisite is sufficient resolution to
-distinguish the objects from each other.
-
-[Illustration: Figure 4—A relation of objects—displayed and simulated]
-
-
-MATHEMATICAL MODEL
-
-The mathematical model must represent both the environment and the SOM
-and for reasons given in the companion paper each is represented as a
-metrizable topology. For uniqueness we factor each space into equal
-parts and represent the environment as the channel
-
-W ⟶ X. (Ref. 10a)
-
-Consider now the SOM to be represented by the cascaded channels
-
-X ⟶ Y ⟶ Z
-
-where X ⟶ Y is a variable which represents the reorganization of the
-SOM existing input-output relation represented by Y ⟶ Z.
-
-The solution of the three channels-in-cascade problem
-
-W ⟶ X ⟶ Y ⟶ Z,
-
-where p(W) (11), p(X), p(X|W), p(Y), p(Z), p(Z|Y) are fixed, yields
-that middle channel p₀(Y|X), from a set of permissible middle channels
-{p(Y|X)}, which maximizes R(Z,W).
-
-Then the resulting middle channel describes that reorganization of the
-SOM which yields the optimum simulation of W ⟶ X by the SOM, within the
-constraints upon Ch(Z,Y).
-
-The solution (the middle channel) depends of course on the particular
-end channels. Obviously the algorithm which is used to find the
-solution does not. It follows that if some physical process were
-constrained to carrying out the steps specified by the algorithm,
-said process would be capable of simulation and would exhibit
-self-organization.
-
-Although the formal solution to the three-channels-in-cascade problem
-is not complete, the solution is sufficiently well characterized to
-permit proceeding with a mechanization of the algorithm. A considerable
-portion of the solution is concerned with the decomposition and
-metrization of channels and it is upon this feature that we now focus
-attention.
-
-As suggested in the companion paper, if the dimensionality of the
-spaces is greater than one, the SOM has only one method available (12).
-Consider the decomposition of a space without, for the moment, making
-the distinction between input and output.
-
-Figure 5 depicts objects represented by a (perhaps multidimensional)
-“cloud” of points. In the absence of a preassigned coordinate system,
-the SOM computes the center of gravity of the cloud (which can be
-done in any coordinate system) and describes the points in terms of
-the distance from this center of gravity; or, which is the same, as
-concentric spheres with origin at the center of gravity.
-
-[Illustration: Figure 5—Nilpotent decomposition of a three-dimensional
-space]
-
-The direction of particular point cannot be specified for there is no
-reference radius vector. Since the SOM wants to end up with a cartesian
-coordinate system, it must transform the sphere (a two-dimensional
-surface) into a plane (a two-dimensional surface). Unfortunately, a
-sphere is not homeomorphic to a plane; thus the SOM has to decompose
-the sphere into a cartesian product of a hemisphere (12a) and a
-denumerable group. The SOM then can transform the hemisphere into a
-plane. The points projected onto the plane constitute a space of the
-same character as the one with which the SOM started. Thus, it can
-repeat all operations on the plane (a space of one less dimension) by
-finding the center of gravity and the circle upon which the desired
-point is situated. The circle is similarly decomposed into a line times
-a denumerable group. By repeating this operation as many times as the
-space has dimensions, the SOM eventually arrives at a single point and
-has obtained in the process a description of the space. Since this
-procedure can be carried on by the repeated use of one operator, this
-operator is nilpotent and to reflect this fact as well as the use of a
-projection, we have named this a nilpotent projection operator or NPO
-for short.
-
-
-MECHANIZATION OF THE NPO
-
-Analog computer elements were used to simulate one NPO which was
-tested in the experimental configuration shown in Figure 6. The NPO
-operates upon a channel which is artificially generated from the two
-noise generators i₁ and i₂ and the signal generator i₀ (i₀ may also be
-a noise generator). The NPO accepts the inputs labelled X₁ and X₂ and
-provides the three outputs Ξ₁, Ξ₂, and γ. X₁ is the linear combination
-of the outputs of generators i₁ and i₀, similarly X₂ is obtained from
-i₂ and i₀.
-
-[Illustration: Figure 6—Experimental test configuration for the
-simulation of an NPO]
-
-Obviously, i₀ is an important parameter since it represents the memory
-relating the spaces X₁ and X₂. Ξ₁ has the property that the magnitude
-of its projection on i₀ is a maximum while Ξ₂ to the opposite has a
-zero projection on i₀. γ is the detected version of the eigenvalue of
-Ch(X₂,X₁).
-
-In the companion paper it was shown how one can provide a Euclidean
-geometrical representation of the NPO. This representation is shown in
-Figure 7 which shows the vectors i₀, i₁, i₂, X₁, X₂, Ξ₁, Ξ₂, and the
-angles Θ₁, Θ₂, and γ. The length of a vector is given by
-
-|X| = κₓ(2πε)⁻¹ᐟ² ∈ H(X)
-
-and the angle between two vectors by
-
-|Θ(X₁,X₂)|-sin⁻¹ ∈ -R(X₁,X₂).
-
-The three vectors i₀, i₁, i₂ provide an orthogonal coordinate system
-because the corresponding signals are random, _i.e._,
-
- κ
- R(i₀,i₁,i₂) ≡ 0.
-
-As external observers we have a prior knowledge of this coordinate
-system; however, the NPO is given only the vectors X₁ and X₂ in the i₀
-⨉ i₁ and i₀ ⨉ i₂ planes. The NPO can reconstruct the entire geometry
-but the actual output Ξ obviously is constrained to lie in the plane of
-the input vector X. The following formulas are typical of the relations
-present.
-
- |Ξ₁|
- tan β = ————
- |Ξ₂|
-
- cos Θ = cos 2β csc 2γ
-
- cos 2β
- cos 2Θ₁ = -1 + 2 ———————
- 1-cos 2γ
-
- cos Θ = cos Θ₁ cos Θ₂.
-
-[Illustration: Figure 7—Geometry of the NPO]
-
-[Illustration: Figure 8—NPO run number 5]
-
-[Illustration: Figure 9—NPO run number 6]
-
-We have obtained a complete description of the NPO which involves 74
-formulas. These treat the noise in the various outputs, invariances of
-the NPO and other interesting features. A presentation of these would
-be outside of the scope of this paper and would tend to obscure the
-main features of the NPO. Thus, we show here only a typical sample of
-the computer simulation, Figure 8 and Figure 9. Conditions for these
-runs are shown in Table I. Run No. 6 duplicates run No. 5 except for
-the fact that i₁ and i₂ were disabled in run No. 6.
-
-Observe that all our descriptions of the NPO and the space it is to
-decompose have been time invariant while the signals shown in the
-simulation are presented as functions of time. The conversion may be
-effected as follows: Given a measurable (single-valued) function
-
- x = x(t)t ∊ T
- where
- μ(T) > 0
- we define the space
- X = {x = x(t) ∍ t ∊ T}
-
- and a probability distribution
-
- μ(x⁻¹(X′))
- P(X′) = —————————— X′ open ⊂ X
- μ(T)
- on that space.
-
- TABLE I
- Legend for Traces of Figures 8 and 9
- ---------+-------+-------+--------+-----+----------+-------+--------
- Trace | | | | | | |
- Number | 1 | 2 | 3 | 4 | 5 | 6 | 7
- ---------+-------+-------+--------+-----+----------+-------+--------
- Symbol | X₂ | X₁ | γ | β | i | dξ₂/dτ | dξ₁/dτ
- ---------+-------+-------+--------+-----+----------+--------+-------
- run No. 5| | | | | | |
- | | | | | | |
- signal |7½ Vrms|7½ Vrms| π ptop | |35.6 m cps| |
- | | | | | | |
- noise |16 Vrms|15 Vrms| π/9 | | | |
- | | | ptop[5]| |sine wave | |
- | | | | | | |
- DC | 0 | 0 | | | | |
- | | | | | | |
- power s/n| 1/4 | 1/4 | 81/1 | | | 0 | 1/2[6]
- | | | | | | |
- terminal | | | | | | |
- value | | | π/4 | π/4 | | |
- ---------+-------+-------+--------+-----+----------+--------+-------
- run No. 6| | | | | | |
- | | | | | | |
- signal |7½ Vrms|7½ Vrms| π ptop | |35.6 m cps| |
- | | | | | | |
- noise | 0 | 0 | 0[7] | |sine wave | |
- | | | | | | |
- DC | -30V | 0 | | | | |
- | | | | | | |
- power s/n| ∞ | ∞ | ∞ | | | 0 | ∞
- | | | | | | |
- terminal | | | | | | |
- value | | | π/4 | π/4 | | |
- ---------+-------+-------+--------+-----+----------+--------+-------
-
-[5] Observed from Oscillogram
-
-[6] Computed
-
-[7] Observed from Oscillogram
-
-Then (X,p(X)) is a stochastic space in our usual sense and x(T) is a
-stochastic variable. Two immediate consequences are:
-
-P(X) is stationary (P(X) is not a function of t ∊ T), and no question
-of ergodicity arises.
-
-
-NETWORKS OF NPO’S
-
-A network of NPO’s may constitute anything from a SOM to a
-preprogrammed detector, depending upon the relative amount of
-preprogramming included. Two methods of preprogramming are: (1) Feeding
-a signal out of a permanent storage into some of the inputs of the
-network of NPO’s. This a priori copy need not be perfect, because the
-SOM will measure the angles Θᵢ anyhow. (2) Feedback, which, after all,
-is just a way of taking advantage of the storage inherent in any delay
-line. (We implicitly assume that any reasonable physical realization
-of an NPO will include a delay T between the x input and the ξ output
-which is not less than perhaps 10⁻¹ times the time constant of the
-internal feedback loop in the γ computation.)
-
-Simulation of channels that possess a discrete component requires
-feedback path(s) to generate the required free products of the finitely
-generated groups. Then, such a SOM converges to a maximal subgroup of
-the group describing the symmetry of the signal that is a free product
-available to this SOM.
-
-Because a single NPO with 1 ≤ n₀ ≤ K₀ is isomorphic (provides the same
-input to output mapping) to a suitable network of NPO’s with n₀ = 1, it
-suffices to study only networks of NPO’s with n₀ = 1.
-
-Figure 10 is largely self-explanatory. Item a is our schematic symbol
-for a single NPO with n₀ = 1. Items b, d (including larger feedback
-loops), and f are typical of artificial intelligence networks. Item c
-is employed to effect the level changing required in order to apply the
-three channels in cascade algorithm to the solution of one-dimensional
-coding problems. Observe that items c and e are the only configurations
-requiring the γ output. Item d may be used as a limiter by making T⁻¹
-high compared to the highest frequency present in the signal. Observe
-that item e is the only application of NPO’s that requires either the
-ξ₂ or β outputs. Item f serves the purpose of handling higher power
-levels into and out of what effectively is a single (larger) NPO.
-
-[Illustration]
-
-[Illustration: Figure 10—Some possible networks of NPO’s]
-
-
-CONCLUSION
-
-The definition of self-organizing behavior suitably represented has
-permitted the use of Information Theoretic techniques to synthesize
-a (mathematical) mechanism for a self-organizing machine. Physical
-mechanization in the form of an NPO has been accomplished and has
-introduced the experimental phase of the program. From among the many
-items deserving of further study we may mention: more economical
-physical mechanization through introduction of modern technology;
-identification of networks of NPO’s with their group theoretic
-descriptions; analysis of the dimensionality of tasks which a SOM might
-be called on to simulate, and prototype SOM applications to related
-tasks. It is hoped that progress along these lines can be reported in
-the future.
-
-
-
-
-REFERENCES
-
-
- 1. Ścibor-Marchocki, Romuald I.,
- “A Topological Foundation for Self-Organization,”
- Anaheim, California:Northrop Nortronics, NSS Report 2828,
- November 14, 1963
-
- 2. It is true that our definition is very similar to that proposed
- by Hawkins (reference 5). Compare for example his definition of
- learning machines (page 31 of reference 5). But the subsequent
- developments reviewed therein are different from the one we have
- followed.
-
- 3. Ashby, W. R.,
- “The Set Theory of Mechanism and Homeostasis,”
- Technical Report 7, University of Illinois, September 1962
-
- 4. Ashby, W. R.,
- “Systems and Information,”
- _Transactions PTGME_ =MIL-7=:94-97 (April-July, 1963)
-
- 5. Hawkins, J. K.,
- “Self-Organizing Systems—A Review and Commentary,”
- _Proc. IRE_. =49=:31-48 (January 1961)
-
- 6. Mesarovic, M. D.,
- “On Self Organizational Systems,”
- Spartan Books, pp. 9-36, 1962
-
- 7. Braverman, D.,
- “Learning Filters for Optimum Pattern Recognition,”
- _PGIT_ =IT-8=:280-285 (July 1962)
-
- 8. We make the latter statement despite the fact that we employ a
- statistical treatment of self-organization. We may predict the
- performance of, for example, the NPO by using a statistical
- description, but it does not necessarily follow that the NPO
- computes statistics.
-
- 9. McCulloch, W. S., and Pitts, W.,
- “A Logical Calculus of the Ideas Imminent in Nervous Activity,”
- _Bull-Math. Biophys_ =5=:115 (1943)
-
- 10. Newell, A., Shaw, J. C., and Simon, H. A.,
- “Empirical Explorations of the Logic Theory Machine:
- A Case Study in Heuristic,”
- _Proc. WJCC_, pp. 218-230, 1957
-
- 10a. The spaces W, X, Y, and Z are stochastic spaces; that is,
- each space is defined as the ordered pair (X,p(X)) where
- p(X) = {p(x) ∋ x ∈ X}, p(x) ≥ 0, x ∈ X and ∫x p(x)dx = 1.
- Such spaces possess a metrizable topology.
-
- 11. We use the following convention for probability distributions:
- if the arguments of p( ) are different, they are different
- functions, thus: p(x) ≠ p(y) even if y = x.
-
- 12. One can prove the existence of a metric directly but in order
- to perform the metrization the space has to be decomposed first.
- But decomposing a space without having a metric calls for a neat
- trick, accomplished (as far as we know) only by the method used
- by the SOM.
-
- 12a. In this example we use a hemisphere; in general, it would be
- a spherical cap.
-
-
-
-
-A Topological Foundation for Self-Organization
-
-
- R. I. ŚCIBOR-MARCHOCKI
-
- _Northrop Nortronics_
- _Systems Support Department_
- _Anaheim, California_
-
- It is shown that by the use of Information
- Theory, any metrizable topology may be metrized
- as an orthogonal Euclidean space (with a random
- Gaussian probability distribution) times
- a denumerable random cartesian product of
- irreducible (wrt direct product) denumerable
- groups. The necessary algorithm to accomplish
- this metrization from a statistical basis is
- presented. If such a basis is unavailable,
- a certain nilpotent projection operator has
- to be used instead, as is shown in detail in
- the companion paper. This operator possesses
- self-organizing features.
-
-
-INTRODUCTION
-
-In the companion article[8] we will define a self-organizing system
-as one which, after observing the input and output of an unknown
-phenomenon (transfer relation), organizes itself into a simulation of
-the unknown phenomenon.
-
-[8] Kleyn, P. A., “Conceptual Design of Self-Organizing Machines,”
-Anaheim, California:Northrop Nortronics, NSS Report 2832, Nov. 14, 1963.
-
-Within the mathematical model, the aforementioned phenomenon may be
-represented as a topological space thus omitting for the moment the
-(arbitrary) designation of input and output which, as will be shown,
-bears on the question of uniqueness. Hence, for the purpose of this
-paper, which emphasizes the mathematical foundation, an intelligent
-device is taken as one which carries out the task of studying a space
-and describing it.
-
-In keeping with the policy that one should not ask someone (or
-something) else to do a task that he could not do himself (at least in
-principle), let us consider how we would approach such a problem.
-
-In the first place, we have to select the space in which the problem is
-to be set. The most general space that we feel capable of tackling is
-a metrizable topology. On the other hand, anything less general would
-be unnecessarily restrictive. Thus, we choose a metrizable topological
-space.
-
-As soon as we have made this choice, we regret it. In order to
-improve the situation somewhat, we show that there is no (additional)
-loss of generality in using an orthogonal Euclidean space times[9]
-a denumerable random cartesian product of irreducible (wrt direct
-product) denumerable groups.
-
-This paper provides a survey of the problem and a method for solving
-it which is conceptually clear but not very practical. The companion
-paper[10] provides a practical method for solving this problem by means
-of the successive use of a certain nilpotent projection operator.
-
-[9] Random cartesian product.
-
-[10] Kleyn, P. A., “Conceptual Design of Self-Organizing Machines,”
-Anaheim, California:Northrop Nortronics, NSS Report 2832, Nov. 14, 1963.
-
-METRIZATION
-
-We start with a metrizable topological space. There are many equivalent
-axiomatizations of a metrizable topology; _e.g._, see Kelley. Perhaps
-the easiest way to visualize a metrizable topology is to consider that
-one was given a metric space but that he lost his notes in which the
-exact form of the metric was written down. Thus one knows that he can
-do everything that he could in a metric space, if only he can figure
-out how.
-
-The “figuring out how” is by no means trivial. Here, it will be assumed
-that a cumulative probability distribution has been obtained on the
-space by one of the standard methods; bird in cage,[11] Munroe I,[12]
-Munroe II,[13] ordering (see Halmos[14] or Kelley[15]). This cumulative
-probability distribution is a function on X onto the interval [0,1] of
-real numbers. The inverse of this function, which exists by the Radon
-Nikodym theorem, provides a mapping from the real interval onto the
-non-trivial portion of X. This mapping induces all of the pleasant
-properties of the real numbers on the space X: topological, metric, and
-ordering.
-
-Actually, it turns out that, especially if the dimensionality of
-the space is greater than one, the foregoing procedure not only
-provides one metrization, but many. Indeed, this lack of uniqueness
-is what makes the procedure exceedingly difficult. Only by imposing
-some additional conditions that result in the existence of a unique
-solution, does the problem become tractable.
-
-We choose to impose the additional condition that the resulting metric
-space be a Euclidean geometry with a rectangular coordinate system.
-
-[11] Harman, W. W., “Principles of the Statistical Theory of
-Communication,” New York, New York:McGraw-Hill, 1963.
-
-[12] Munroe, M. E., “Introduction to Measure and Integration,”
-Cambridge, Mass.:Addison-Wesley, 1953.
-
-[13] Munroe, M. E., “Introduction to Measure and Integration,”
-Cambridge, Mass.:Addison-Wesley, 1953.
-
-[14] Halmos, P. R., “Measure Theory,” Princeton, New Jersey:D. Van
-Nostrand Co., Inc., 1950.
-
-[15] Kelley, J. L., “General Topology,” Princeton, New Jersey:D. Van
-Nostrand Co., Inc., 1955.
-
-Even this always does not yield uniqueness, but we will show the
-additional restriction that will guarantee uniqueness after the
-necessary language is developed. Since all metrizations of a given
-metrizable topology are isomorphic, in the quotient class the
-orthogonal Euclidean geometry serves the purpose of being a convenient
-representative of the unique element resulting from a given metrizable
-topology.
-
-Furthermore, the same comment applies to the use of a Gaussian
-distribution as the probability distribution on this orthogonal
-Euclidean geometry. Namely, the random Gaussian distribution on an
-orthogonal Euclidean geometry is a convenient representative member of
-the equivalence class which maps into one element (stochastic space) of
-the quotient class.
-
-
-Information Theory
-
-Now, we will show that Information Theory provides the language
-necessary to describe the metrization procedure in detail.
-
-It is possible to introduce Information Theory axiomatically by a
-suitable generalization of the axioms[16] in Feinstein.[17] But
-to simplify the discussion here, we will use the less elegant but
-equivalent method of defining certain definite integrals. The
-probability density distribution p is defined from the cumulative
-probability distribution P by
-
- P(X′) = ∫X′_{measurable ⊂ X} p(x)dx. (1)
-
- Then the information rate H is defined as
-
- H(X) = -∫ₓp(x) ln κ p(x)dx (2)
-
- where kappa has (carries) the units of X. Finally,
- the channel rate R is defined as
-
- R(⨀Xᵢ) = ΣH(Xᵢ) - H(X), (3)
- I I
-
- where X is the denumerable[18] cartesian product space
-
- X = ⨂Xᵢ. (4)
- I
-
-[16] Feinstein uses his axioms only in finite space X; _i.e._, card(X)
-< K₀.
-
-[17] Feinstein, A., “Foundations of Information Theory,” New York, New
-York: McGraw-Hill, 1958.
-
-[18] If I is infinite, certain precautions have to be exercised.
-
-Next, we define the angle Θ
-
- |Θ(⨀Xᵢ)| = sin⁻¹_e_^{-R(⨀Xᵢ)} (5)
- I
- and the norm
-
- |X| = κ(2π_e_)⁻¹ᐟ² _e_^{(HX)}. (6)
-
- Now, if[19] a statistically independent basis; _i.e._, one for which
- κ
- R(⨀Xᵢ) ≡ constant, (7)
- I
-
-can be provided in terms of one-dimensional components; _i.e._, none
-of them can be decomposed further, then it is just the usual problem
-of diagonalization of a symmetric matrix by means of a congruence
-transformation to provide an orthogonal coordinate system. Furthermore,
-for uniqueness, we arrange the spectrum in decreasing order. Then,
-by means of the Radon Nikodym theorem applied to each of these
-one-dimensional axes, the probability distribution may be made; _e.g._,
-Gaussian, if desired. Thus, we obtain the promised orthogonal Euclidean
-space.
-
-[19] This “if” is the catch that makes all methods of metrization of a
-space of dimensionality higher than one impractical, except the method
-of successive projections upon unit spheres centered at the center of
-gravity. The method of using that nilpotent projection operator is
-described in the companion paper(see footnote page 65).
-
-
-Channel
-
-At this time we can state the remaining additional condition required
-that a decomposition be unique. The index space I has to be partitioned
-into exactly two parts, say I′ and I″; _i.e._,
-
- I′ ∪ I″ = I (8)
-
- I′ ∩ I″ = φ,
-
- such that
-
- dim(X′) = dim(X″), (9)
-
- where
-
- X′ = ⨂Xᵢ (10)
- I′
-
- X″ = ⨂Xᵢ.
- I″
-
-(If dim (X) is odd, then we have to cheat a little by putting in an
-extra random dummy dimension.) And then the decomposition of the space
-
-X = ⨂Xᵢ (11) I
-
-has to be carried out so that this partitioning is preserved.
-Since this partitioning is arbitrary (as far as the mathematics is
-concerned), it is obvious that a space which is not partitioned will
-have many (equivalent) decompositions. On the other hand, if the
-partitioning is into more than two parts, then the existence of a
-decomposition is not guaranteed.
-
-A slight penalty has to be paid for the use of this partitioning,
-namely: instead of eventually obtaining a random cartesian product of
-one-dimensional spaces, we obtain an extended channel (with random
-input) of single-dimensional channels. It is obvious that if we were
-to drop the partitioning temporarily, each such single-dimensional
-channel would be further decomposed into two random components. This
-decomposition is not unique. But one of these equivalent decompositions
-is particularly convenient; namely, that decomposition where we take
-the component out of the original X′ and that which is random to it,
-say V. This V (as well as the cartesian product of all such V’s,
-which of necessity are random) is called the linearly additive noise.
-The name “linearly additive” is justified because it is just the
-statistical concept isomorphic to the linear addition of vectors in
-orthogonal Euclidean geometry. (The proof of this last statement is not
-completed as yet.)
-
-
-Denumerable Space
-
-The procedure for this decomposition was worded to de-emphasize the
-possible presence of a denumerable (component of the) space. Such a
-component may be given outright; otherwise, it results if the space was
-not simply connected. Any denumerable space is zero dimensional, as may
-be verified easily from the full information theoretic definition of
-dimensionality.
-
-The obvious way of disposing of a denumerable space is to use the
-conventional mapping that converts a Stieltjes to a Lebesque integral,
-using fixed length segments. (It can be shown that H is invariant
-under such a mapping.) Unfortunately, while this mapping followed by
-a repetition of the preceding procedure will always solve a given
-problem (no new[20] denumerable component _need_ be generated on the
-second pass), little insight is provided into the structure of the
-resulting space. On the other hand, because channels under cascading
-constitute a group, any such denumerable space is a representation of a
-denumerable group.
-
-[20] Only non-cyclic irreducible (wrt direct product) denumerable group
-components of the old denumerable space will remain.
-
-
-SUMMARY
-
-In summary, the original metrizable topological space was decomposed
-into an orthogonal Euclidean space times[21] a denumerable random
-cartesian product of irreducible (wrt direct product) denumerable
-groups. Thus, since any individual component of a random cartesian
-product may be studied independently of the others, all that one needs
-to study is: (1) a Gaussian distribution on a single real axis and (2)
-the irreducible denumerable groups.
-
-[21] Random cartesian product.
-
-Finally, it should be emphasized that there are only these two ways
-of decomposing a metrizable topology; (1) if a (statistical) basis
-is given, use the diagonalization of a symmetric matrix algorithm
-described earlier (and given in detail in the three channels in cascade
-problem), and (2) otherwise use a suitable network of the NPO’s with
-n₀=1. Of course, any hybrid of these two methods may be employed as
-well.
-
-
-
-
-On Functional Neuron Modeling
-
-
- C. E. HENDRIX
-
- _Space-General Corporation_
- _El Monte, California_
-
-There are two very compelling reasons why mathematical and physical
-models of the neuron should be built. Model building, while widely
-used in the physical sciences, has been largely neglected in biology.
-However, there can be little doubt that building neuron models
-will increase our understanding of the function of real neurons,
-if experience in the physical sciences is any guide. Secondly,
-neuron models are extremely interesting in their own right as new
-technological devices. Hence, the interest in, and the reason for
-symposia on self-organizing systems.
-
-We should turn our attention to the properties of real neurons, and
-see which of them are the most important ones for us to imitate.
-Obviously, we cannot hope to imitate _all_ the properties of a living
-neuron, since that would require a complete simulation of a living,
-metabolizing cell, and a highly specialized one at that; but we
-can select those functional properties which we feel are the most
-important, and then try to simulate those.
-
-The most dramatic aspect of neuron function is, of course, the axon
-discharge. It is this which gives the neuron its “all-or-nothing”
-character, and it is this which provides it with a means for
-propagating its output pulses over a distance. Hodgkin and Huxley (1)
-have developed a very complete description of this action. Their model
-is certainly without peer in describing the nature of the real neuron.
-
-On the technological side, Cranes’ “neuristors” (2) represent a class
-of devices which imitate the axonal discharge in a gross sort of way,
-without all the subtle nuances of the Hodgkin-Huxley model. Crane has
-shown that neuristors can be combined to yield the various Boolean
-functions needed in a computer.
-
-However, interesting as such models of the axon are, there is some
-question as to their importance in the development of self-organizing
-systems. The pulse generation, “all-or-nothing” part of the axon
-behavior could just as well be simulated by a “one-shot” trigger
-circuit. The transmission characteristic of the axon is, after all,
-only Nature’s way of sending a signal from here to there. It is
-an admirable solution to the problem, when one considers that it
-evolved, and still works, in a bath of salt water. There seems little
-point, however, in a hardware designer limiting himself in this way,
-especially if he has an adequate supply of insulated copper wire.
-
-If the transmission characteristic of the axon is deleted, the
-properties of the neuron which seem to be the most important in the
-synthesis of self-organizing systems are:
-
- a. The neuron responds to a stimulus with an electrical
- pulse of standard size and shape. If the stimulus
- continues, the pulses occur at regular intervals
- with the rate of occurrence dependent on the
- intensity of stimulation.
-
- b. There is a threshold of stimulation. If the
- intensity of the stimulus is below this threshold,
- the neuron does not fire.
-
- c. The neuron is capable of temporal and spatial
- integration. Many subthreshold stimuli arriving at
- the neuron from different sources, or at slightly
- different times, can add up to a sufficient level to
- fire the neuron.
-
- d. Some inputs are excitatory, some are inhibitory.
-
- e. There is a refractory period. Once fired, there is
- a subsequent period during which the neuron cannot
- be fired again, no matter how large the stimulus.
- This places an upper limit on the pulse rate of any
- particular neuron.
-
- f. The neuron can learn. This property is conjectural
- in living neurons, since it appears that at
- the present time learning has not been clearly
- demonstrated in isolated living neurons.
- However, the learning property is basic to all
- self-organizing models.
-
-Neuron models with the above characteristics have been built, although
-none seem to have incorporated _all_ of them in a single model. Harman
-(3) at Bell Labs has built neuron models which have the characteristics
-(a) through (e), with which he has built extremely interesting devices
-which simulate portions of the peripheral neuron system.
-
-Various attempts at learning elements have been made, perhaps best
-exemplified by those of Widrow (4). These devices are capable of
-“learning,” but are static, and lack all the temporal characteristics
-listed in (a) through (e). Such devices can be used to deal with
-temporal patterns only by a mapping technique, in which a temporal
-pattern is converted to a spatial one.
-
-Having listed which seem to be the important properties of a neuron, it
-is possible to synthesize a simple model which has all of them.
-
-A number of input stimuli are fed to the neuron through a resistive
-summing network which establishes the threshold and accomplishes
-spatial integration. The voltage at the summing junction triggers a
-“one-shot” circuit, which, by its very nature, accomplishes pulse
-generation and exhibits temporal integration and a refractory period.
-The polarity of an individual input determines whether it shall be
-excitatory or inhibitory. This much of the circuitry is very similar to
-Harmon’s model.
-
-Learning is postulated to take place in the following way: when the
-neuron fires, an outside influence (the environment, or a “trainer”)
-determines whether or not the result of firing was desirable or not.
-If it was desirable, the threshold of the neuron is lowered, making
-it easier to fire the next time. If the result was not desirable, the
-threshold is raised, making it more difficult for the neuron to fire
-the next time.
-
-In a self-organizing system, many model neurons would be
-interconnected. A “punish-reward” (P-R) signal would be connected to
-all neurons in common. However, means would be provided for only those
-which have recently fired to be susceptible to the effects of the P-R
-signal. Therefore, only those which had taken part in a recent response
-are modified. This idea is due to Stewart (5), who applies it to his
-electrochemical devices instead of to an electronic device.
-
-The mechanization of the circuitry is rather straight-forward. A
-portion of the output of the pulse generator is routed through a
-“pulse-stretcher” or short-term memory which temporarily records the
-fact that the neuron has recently fired. The pulse-stretcher output
-controls a gate, which either accepts or rejects the P-R signal. The
-P-R signal can take on only three values, a positive level, zero, or
-a negative level, depending on whether the signal is “punish,” “no
-action,” or “reward.” Finally, the gate output controls a variable
-resistor, which is part of the resistive summing network. Figure 1 is a
-block diagram of the complete model.
-
-Note that this device differs from the usual “Perceptron” configuration
-in that the threshold resistor is the only variable element, instead
-of having each input resistor a variable weighting element. This
-simplification could lead to a situation where, to prepare a specified
-task, more single-variable neurons would be required than would
-multivariable ones. This possible disadvantage is partially, at least,
-offset by the very simple control algorithm which is contained in the
-design of the model, and is not the matter of great concern which it
-seems to be for most multivariable models.
-
-[Illustration: Figure 1—Block diagram of neuron model]
-
-Hand simulations of the action of this type of model suggest that a
-certain amount of randomness would be desirable. It appears that a
-self-organizing system built of these elements, and of sufficient
-complexity to be interesting, would have a fair number of recirculating
-loops, so that spontaneous activity would be maintained in the absence
-of input stimulus. If this is the case, then randomness could easily
-be introduced by adding a small amount of noise from a random noise
-generator to the signal on the P-R bus. Thus, any neurons which
-spontaneously fire would be continually having their thresholds
-modified.
-
-The mechanization of the model is not particularly complex, and can
-be estimated as follows: The one-shot pulse generator would require
-two transistors, the pulse stretcher one more. The bi-directional gate
-would require a transistor and at least two diodes.
-
-Several candidates for the electrically-controllable variable
-resistor are available (6). Particularly good candidates appear to
-be the “Memistor” or plating cell developed by Widrow (7), the solid
-state version of it by Vendelin (8), and the “solion” (9). All are
-electrochemical devices in which the resistance between two terminals
-is controlled by the net charge flow through a third terminal. All are
-adaptable to this particular circuit.
-
-Of the three, however, the solion appears at first glance to have the
-most promise in that its resistance is of the order of a few thousand
-ohms (rather than the few ohms of the plating cells) which is more
-compatible with ordinary solid-state circuitry. Solions have the
-disadvantage that they can stand only very low voltages (less than 1
-volt) and in their present form require extra bias potentials. If these
-difficulties can be overcome, they offer considerable promise.
-
-In summary, it appears that a rather simple neuron model can be built
-which can mimic most of the important functions of real neurons. A
-system built of these could be punished or rewarded by an observer,
-so that it could be trained to give specified responses to specified
-stimuli. In some cases, the observer could be simply the environment,
-so that the system would learn directly from experience, and would be
-therefore a self-organizing system.
-
-
-REFERENCES
-
- 1. Hodgkin, A. L., and Huxley, A. L.,
- “A Quantitative Description of Membrane Current and its
- Application to Conduction and Excitation in Nerve,”
- _J. Physiol._ =117=:500-544 (August 1952)
-
- 2. Crane, H. D.,
- “Neuristor—A Novel Device and System Concept,”
- _Proc. IRE_ =50=:2048-2060 (Oct. 1962)
-
- 3. Harmon, L. D., Levinson, J., and Van Bergeijk, W. A.,
- “Analog Models of Neural Mechanism,”
- _IRE Trans. on Information Theory_ =IT-8=:107-112
- (Feb. 1962)
-
- 4. Widrow, B., and Hoff, M. E.,
- “Adaptive Switching Circuits,”
- Stanford Electronics Lab Tech Report 1553-1, June 1960
-
- 5. Stewart, R. M.,
- “Electrochemical Wave Interactions and Extensive Field Effects
- in Excitable Cellular Structures,”
- First Pasadena Invitational Symposium on Self-Organizing Systems,
- Calif. Institute of Technology, Pasadena, Calif., 14 Nov. 1963
-
- 6. Nagy, G.,
- “A Survey of Analog Memory Devices,”
- _IEEE Trans. on Electronic Cmptrs._ EC-12:388-393 (Aug. 1963)
-
- 7. Widrow, B.,
- “An Adaptive Adaline Neuron Using Chemical Memistors,”
- Stanford Electronics Lab Tech Report 1553-2, Oct. 1960
-
- 8. Vendelin, G. D.,
- “A Solid State Adaptive Component,”
- Stanford Electronics Lab Tech Report 1853-1, Jan. 1963
-
- 9. “Solion Principles of Electrochemistry and Low-Power
- Electrochemical Devices,”
- Dept. of Comm., Office of Tech. Serv. =PB= 131931
- (U. S. Naval Ord. Lab., Silver Spring, Md., Aug. 1958)
-
-
-
-
-Selection of Parameters for Neural Net Simulations[22]
-
-
- R. K. OVERTON
-
- _Autonetics Research Center_
- _Anaheim, California_
-
-Research of high quality has been presented at this Symposium. Of
-particular interest to me were the reports of the Aeronutronic group
-and the Librascope group. The Aeronutronic group was commendably
-systematic in its investigations of different arrangements of linear
-threshold elements, and the Librascope data, presenting the effects of
-attaching different values to the parameters of simulated neurons, are
-both systematic and interesting.
-
-Unfortunately, however, interest in such research can obscure a more
-fundamental question which seems to merit study. That question concerns
-the parameters, or attributes, which describe the simulated neuron.
-Specifically, which parameters or attributes should be selected for
-simulation? (For example, should a period of supernormal sensitivity be
-simulated following an absolutely refractory period?)
-
-Some selection obviously has to be made. Librascope, which is
-trying to simulate neurons more or less faithfully, plans to build
-a net of ten simulated neurons. In contrast, General Dynamics/Fort
-Worth, with roughly the same degree of effort, is working with 3900
-unfaithfully-simulated neurons. This comparison is not a criticism
-of either group; the Librascope team has simply selected many more
-parameters for simulation than has the General Dynamics group. Each
-can make the selections it prefers, because the parameters of real
-neurons which are necessary and sufficient for learning have not been
-exhaustively identified.
-
-From the point of view of one whose interests include real neurons,
-this lack of identification is unfortunate. I once wrote a book which
-included some guesses about the essential attributes of neurons. Since
-that time, many neuron simulation programs have been written. But these
-programs, although interesting and worthwhile in their own right, have
-done little to answer the question of the necessary parameters. That
-is, they do not make much better guesses possible. And yet better
-guesses would also make for more “intelligent” machines.
-
-[22] This paper, submitted after the Symposium, represents a more
-detailed presentation of some of the issues raised in the discussion
-sessions at the Symposium and hence, constitutes a worthwhile addition
-to the Proceedings.
-
-
-
-
-INDEX OF INVITED PARTICIPANTS
-
-
- MICHAEL ARBIB Massachusetts Institute of Technology
- ROBERT H. ASENDORF Hughes Research Laboratories/ Malibu
- J. A. DALY Astropower/Newport Beach
- GEORGE DeFLORIO System Development Corp./Santa Monica
- DEREK H. FENDER California Institute of Technology
- LEONARD FRIEDMAN Space Technology Labs./Redondo Beach
- JAMES EMMETT GARVEY ONR/Pasadena
- THOMAS L. GRETTENBERG California Institute of Technology
- HAROLD HAMILTON Librascope/Glendale
- JOSEPH HAWKINS Aeronutronic/Newport Beach
- CHARLES HENDRIX Space-General Corp./El Monte
- R. D. JOSEPH Astropower/Newport Beach
- PETER A. KLEYN Nortronics/Anaheim
- JOHN KUHN Space-General Corp./El Monte
- FRANK LEHAN Space-General Corp./El Monte
- EDWIN LEWIS Librascope/Glendale
- PETER C. LOCKEMANN California Institute of Technology
- GILBERT D. McCANN California Institute of Technology
- C. J. MUNCIE Aeronutronic/Newport Beach
- C. OVERMIER Nortronics/Anaheim
- RICHARD K. OVERTON Autonetics/Anaheim
- DIANE RAMSEY Astropower/Newport Beach
- RICHARD REISS Librascope/Glendale
- R. I. ŚCIBOR-MARCHOCKI Nortronics/Anaheim
- JAMES J. SPILKER Philco/Palo Alto
- ROBERT M. STEWART Space-General Corp./El Monte
- HENNIG STIEVE California Institute of Technology
- RICHARD TEW Space-General Corp./El Monte
- JOHN THORSEN University of California/Los Angeles
- RICHARD VINETZ Librascope/Glendale
- CHRISTOPH von CAMPENHAUSEN California Institute of Technology
- DAVID VOWLES California Institute of Technology
- HORST WOLF Astropower/Newport Beach
-
- U.S. GOVERNMENT PRINTING OFFICE: 1966 O—205-502
-
-*** END OF THE PROJECT GUTENBERG EBOOK SELF-ORGANIZING SYSTEMS, 1963 ***
-
-Updated editions will replace the previous one--the old editions will
-be renamed.
-
-Creating the works from print editions not protected by U.S. copyright
-law means that no one owns a United States copyright in these works,
-so the Foundation (and you!) can copy and distribute it in the
-United States without permission and without paying copyright
-royalties. Special rules, set forth in the General Terms of Use part
-of this license, apply to copying and distributing Project
-Gutenberg-tm electronic works to protect the PROJECT GUTENBERG-tm
-concept and trademark. Project Gutenberg is a registered trademark,
-and may not be used if you charge for an eBook, except by following
-the terms of the trademark license, including paying royalties for use
-of the Project Gutenberg trademark. If you do not charge anything for
-copies of this eBook, complying with the trademark license is very
-easy. You may use this eBook for nearly any purpose such as creation
-of derivative works, reports, performances and research. Project
-Gutenberg eBooks may be modified and printed and given away--you may
-do practically ANYTHING in the United States with eBooks not protected
-by U.S. copyright law. Redistribution is subject to the trademark
-license, especially commercial redistribution.
-
-START: FULL LICENSE
-
-THE FULL PROJECT GUTENBERG LICENSE
-PLEASE READ THIS BEFORE YOU DISTRIBUTE OR USE THIS WORK
-
-To protect the Project Gutenberg-tm mission of promoting the free
-distribution of electronic works, by using or distributing this work
-(or any other work associated in any way with the phrase "Project
-Gutenberg"), you agree to comply with all the terms of the Full
-Project Gutenberg-tm License available with this file or online at
-www.gutenberg.org/license.
-
-Section 1. General Terms of Use and Redistributing Project
-Gutenberg-tm electronic works
-
-1.A. By reading or using any part of this Project Gutenberg-tm
-electronic work, you indicate that you have read, understand, agree to
-and accept all the terms of this license and intellectual property
-(trademark/copyright) agreement. If you do not agree to abide by all
-the terms of this agreement, you must cease using and return or
-destroy all copies of Project Gutenberg-tm electronic works in your
-possession. If you paid a fee for obtaining a copy of or access to a
-Project Gutenberg-tm electronic work and you do not agree to be bound
-by the terms of this agreement, you may obtain a refund from the
-person or entity to whom you paid the fee as set forth in paragraph
-1.E.8.
-
-1.B. "Project Gutenberg" is a registered trademark. It may only be
-used on or associated in any way with an electronic work by people who
-agree to be bound by the terms of this agreement. There are a few
-things that you can do with most Project Gutenberg-tm electronic works
-even without complying with the full terms of this agreement. See
-paragraph 1.C below. There are a lot of things you can do with Project
-Gutenberg-tm electronic works if you follow the terms of this
-agreement and help preserve free future access to Project Gutenberg-tm
-electronic works. See paragraph 1.E below.
-
-1.C. The Project Gutenberg Literary Archive Foundation ("the
-Foundation" or PGLAF), owns a compilation copyright in the collection
-of Project Gutenberg-tm electronic works. Nearly all the individual
-works in the collection are in the public domain in the United
-States. If an individual work is unprotected by copyright law in the
-United States and you are located in the United States, we do not
-claim a right to prevent you from copying, distributing, performing,
-displaying or creating derivative works based on the work as long as
-all references to Project Gutenberg are removed. Of course, we hope
-that you will support the Project Gutenberg-tm mission of promoting
-free access to electronic works by freely sharing Project Gutenberg-tm
-works in compliance with the terms of this agreement for keeping the
-Project Gutenberg-tm name associated with the work. You can easily
-comply with the terms of this agreement by keeping this work in the
-same format with its attached full Project Gutenberg-tm License when
-you share it without charge with others.
-
-1.D. The copyright laws of the place where you are located also govern
-what you can do with this work. Copyright laws in most countries are
-in a constant state of change. If you are outside the United States,
-check the laws of your country in addition to the terms of this
-agreement before downloading, copying, displaying, performing,
-distributing or creating derivative works based on this work or any
-other Project Gutenberg-tm work. The Foundation makes no
-representations concerning the copyright status of any work in any
-country other than the United States.
-
-1.E. Unless you have removed all references to Project Gutenberg:
-
-1.E.1. The following sentence, with active links to, or other
-immediate access to, the full Project Gutenberg-tm License must appear
-prominently whenever any copy of a Project Gutenberg-tm work (any work
-on which the phrase "Project Gutenberg" appears, or with which the
-phrase "Project Gutenberg" is associated) is accessed, displayed,
-performed, viewed, copied or distributed:
-
- This eBook is for the use of anyone anywhere in the United States and
- most other parts of the world at no cost and with almost no
- restrictions whatsoever. You may copy it, give it away or re-use it
- under the terms of the Project Gutenberg License included with this
- eBook or online at www.gutenberg.org. If you are not located in the
- United States, you will have to check the laws of the country where
- you are located before using this eBook.
-
-1.E.2. If an individual Project Gutenberg-tm electronic work is
-derived from texts not protected by U.S. copyright law (does not
-contain a notice indicating that it is posted with permission of the
-copyright holder), the work can be copied and distributed to anyone in
-the United States without paying any fees or charges. If you are
-redistributing or providing access to a work with the phrase "Project
-Gutenberg" associated with or appearing on the work, you must comply
-either with the requirements of paragraphs 1.E.1 through 1.E.7 or
-obtain permission for the use of the work and the Project Gutenberg-tm
-trademark as set forth in paragraphs 1.E.8 or 1.E.9.
-
-1.E.3. If an individual Project Gutenberg-tm electronic work is posted
-with the permission of the copyright holder, your use and distribution
-must comply with both paragraphs 1.E.1 through 1.E.7 and any
-additional terms imposed by the copyright holder. Additional terms
-will be linked to the Project Gutenberg-tm License for all works
-posted with the permission of the copyright holder found at the
-beginning of this work.
-
-1.E.4. Do not unlink or detach or remove the full Project Gutenberg-tm
-License terms from this work, or any files containing a part of this
-work or any other work associated with Project Gutenberg-tm.
-
-1.E.5. Do not copy, display, perform, distribute or redistribute this
-electronic work, or any part of this electronic work, without
-prominently displaying the sentence set forth in paragraph 1.E.1 with
-active links or immediate access to the full terms of the Project
-Gutenberg-tm License.
-
-1.E.6. You may convert to and distribute this work in any binary,
-compressed, marked up, nonproprietary or proprietary form, including
-any word processing or hypertext form. However, if you provide access
-to or distribute copies of a Project Gutenberg-tm work in a format
-other than "Plain Vanilla ASCII" or other format used in the official
-version posted on the official Project Gutenberg-tm website
-(www.gutenberg.org), you must, at no additional cost, fee or expense
-to the user, provide a copy, a means of exporting a copy, or a means
-of obtaining a copy upon request, of the work in its original "Plain
-Vanilla ASCII" or other form. Any alternate format must include the
-full Project Gutenberg-tm License as specified in paragraph 1.E.1.
-
-1.E.7. Do not charge a fee for access to, viewing, displaying,
-performing, copying or distributing any Project Gutenberg-tm works
-unless you comply with paragraph 1.E.8 or 1.E.9.
-
-1.E.8. You may charge a reasonable fee for copies of or providing
-access to or distributing Project Gutenberg-tm electronic works
-provided that:
-
-* You pay a royalty fee of 20% of the gross profits you derive from
- the use of Project Gutenberg-tm works calculated using the method
- you already use to calculate your applicable taxes. The fee is owed
- to the owner of the Project Gutenberg-tm trademark, but he has
- agreed to donate royalties under this paragraph to the Project
- Gutenberg Literary Archive Foundation. Royalty payments must be paid
- within 60 days following each date on which you prepare (or are
- legally required to prepare) your periodic tax returns. Royalty
- payments should be clearly marked as such and sent to the Project
- Gutenberg Literary Archive Foundation at the address specified in
- Section 4, "Information about donations to the Project Gutenberg
- Literary Archive Foundation."
-
-* You provide a full refund of any money paid by a user who notifies
- you in writing (or by e-mail) within 30 days of receipt that s/he
- does not agree to the terms of the full Project Gutenberg-tm
- License. You must require such a user to return or destroy all
- copies of the works possessed in a physical medium and discontinue
- all use of and all access to other copies of Project Gutenberg-tm
- works.
-
-* You provide, in accordance with paragraph 1.F.3, a full refund of
- any money paid for a work or a replacement copy, if a defect in the
- electronic work is discovered and reported to you within 90 days of
- receipt of the work.
-
-* You comply with all other terms of this agreement for free
- distribution of Project Gutenberg-tm works.
-
-1.E.9. If you wish to charge a fee or distribute a Project
-Gutenberg-tm electronic work or group of works on different terms than
-are set forth in this agreement, you must obtain permission in writing
-from the Project Gutenberg Literary Archive Foundation, the manager of
-the Project Gutenberg-tm trademark. Contact the Foundation as set
-forth in Section 3 below.
-
-1.F.
-
-1.F.1. Project Gutenberg volunteers and employees expend considerable
-effort to identify, do copyright research on, transcribe and proofread
-works not protected by U.S. copyright law in creating the Project
-Gutenberg-tm collection. Despite these efforts, Project Gutenberg-tm
-electronic works, and the medium on which they may be stored, may
-contain "Defects," such as, but not limited to, incomplete, inaccurate
-or corrupt data, transcription errors, a copyright or other
-intellectual property infringement, a defective or damaged disk or
-other medium, a computer virus, or computer codes that damage or
-cannot be read by your equipment.
-
-1.F.2. LIMITED WARRANTY, DISCLAIMER OF DAMAGES - Except for the "Right
-of Replacement or Refund" described in paragraph 1.F.3, the Project
-Gutenberg Literary Archive Foundation, the owner of the Project
-Gutenberg-tm trademark, and any other party distributing a Project
-Gutenberg-tm electronic work under this agreement, disclaim all
-liability to you for damages, costs and expenses, including legal
-fees. YOU AGREE THAT YOU HAVE NO REMEDIES FOR NEGLIGENCE, STRICT
-LIABILITY, BREACH OF WARRANTY OR BREACH OF CONTRACT EXCEPT THOSE
-PROVIDED IN PARAGRAPH 1.F.3. YOU AGREE THAT THE FOUNDATION, THE
-TRADEMARK OWNER, AND ANY DISTRIBUTOR UNDER THIS AGREEMENT WILL NOT BE
-LIABLE TO YOU FOR ACTUAL, DIRECT, INDIRECT, CONSEQUENTIAL, PUNITIVE OR
-INCIDENTAL DAMAGES EVEN IF YOU GIVE NOTICE OF THE POSSIBILITY OF SUCH
-DAMAGE.
-
-1.F.3. LIMITED RIGHT OF REPLACEMENT OR REFUND - If you discover a
-defect in this electronic work within 90 days of receiving it, you can
-receive a refund of the money (if any) you paid for it by sending a
-written explanation to the person you received the work from. If you
-received the work on a physical medium, you must return the medium
-with your written explanation. The person or entity that provided you
-with the defective work may elect to provide a replacement copy in
-lieu of a refund. If you received the work electronically, the person
-or entity providing it to you may choose to give you a second
-opportunity to receive the work electronically in lieu of a refund. If
-the second copy is also defective, you may demand a refund in writing
-without further opportunities to fix the problem.
-
-1.F.4. Except for the limited right of replacement or refund set forth
-in paragraph 1.F.3, this work is provided to you 'AS-IS', WITH NO
-OTHER WARRANTIES OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT
-LIMITED TO WARRANTIES OF MERCHANTABILITY OR FITNESS FOR ANY PURPOSE.
-
-1.F.5. Some states do not allow disclaimers of certain implied
-warranties or the exclusion or limitation of certain types of
-damages. If any disclaimer or limitation set forth in this agreement
-violates the law of the state applicable to this agreement, the
-agreement shall be interpreted to make the maximum disclaimer or
-limitation permitted by the applicable state law. The invalidity or
-unenforceability of any provision of this agreement shall not void the
-remaining provisions.
-
-1.F.6. INDEMNITY - You agree to indemnify and hold the Foundation, the
-trademark owner, any agent or employee of the Foundation, anyone
-providing copies of Project Gutenberg-tm electronic works in
-accordance with this agreement, and any volunteers associated with the
-production, promotion and distribution of Project Gutenberg-tm
-electronic works, harmless from all liability, costs and expenses,
-including legal fees, that arise directly or indirectly from any of
-the following which you do or cause to occur: (a) distribution of this
-or any Project Gutenberg-tm work, (b) alteration, modification, or
-additions or deletions to any Project Gutenberg-tm work, and (c) any
-Defect you cause.
-
-Section 2. Information about the Mission of Project Gutenberg-tm
-
-Project Gutenberg-tm is synonymous with the free distribution of
-electronic works in formats readable by the widest variety of
-computers including obsolete, old, middle-aged and new computers. It
-exists because of the efforts of hundreds of volunteers and donations
-from people in all walks of life.
-
-Volunteers and financial support to provide volunteers with the
-assistance they need are critical to reaching Project Gutenberg-tm's
-goals and ensuring that the Project Gutenberg-tm collection will
-remain freely available for generations to come. In 2001, the Project
-Gutenberg Literary Archive Foundation was created to provide a secure
-and permanent future for Project Gutenberg-tm and future
-generations. To learn more about the Project Gutenberg Literary
-Archive Foundation and how your efforts and donations can help, see
-Sections 3 and 4 and the Foundation information page at
-www.gutenberg.org
-
-Section 3. Information about the Project Gutenberg Literary
-Archive Foundation
-
-The Project Gutenberg Literary Archive Foundation is a non-profit
-501(c)(3) educational corporation organized under the laws of the
-state of Mississippi and granted tax exempt status by the Internal
-Revenue Service. The Foundation's EIN or federal tax identification
-number is 64-6221541. Contributions to the Project Gutenberg Literary
-Archive Foundation are tax deductible to the full extent permitted by
-U.S. federal laws and your state's laws.
-
-The Foundation's business office is located at 809 North 1500 West,
-Salt Lake City, UT 84116, (801) 596-1887. Email contact links and up
-to date contact information can be found at the Foundation's website
-and official page at www.gutenberg.org/contact
-
-Section 4. Information about Donations to the Project Gutenberg
-Literary Archive Foundation
-
-Project Gutenberg-tm depends upon and cannot survive without
-widespread public support and donations to carry out its mission of
-increasing the number of public domain and licensed works that can be
-freely distributed in machine-readable form accessible by the widest
-array of equipment including outdated equipment. Many small donations
-($1 to $5,000) are particularly important to maintaining tax exempt
-status with the IRS.
-
-The Foundation is committed to complying with the laws regulating
-charities and charitable donations in all 50 states of the United
-States. Compliance requirements are not uniform and it takes a
-considerable effort, much paperwork and many fees to meet and keep up
-with these requirements. We do not solicit donations in locations
-where we have not received written confirmation of compliance. To SEND
-DONATIONS or determine the status of compliance for any particular
-state visit www.gutenberg.org/donate
-
-While we cannot and do not solicit contributions from states where we
-have not met the solicitation requirements, we know of no prohibition
-against accepting unsolicited donations from donors in such states who
-approach us with offers to donate.
-
-International donations are gratefully accepted, but we cannot make
-any statements concerning tax treatment of donations received from
-outside the United States. U.S. laws alone swamp our small staff.
-
-Please check the Project Gutenberg web pages for current donation
-methods and addresses. Donations are accepted in a number of other
-ways including checks, online payments and credit card donations. To
-donate, please visit: www.gutenberg.org/donate
-
-Section 5. General Information About Project Gutenberg-tm electronic works
-
-Professor Michael S. Hart was the originator of the Project
-Gutenberg-tm concept of a library of electronic works that could be
-freely shared with anyone. For forty years, he produced and
-distributed Project Gutenberg-tm eBooks with only a loose network of
-volunteer support.
-
-Project Gutenberg-tm eBooks are often created from several printed
-editions, all of which are confirmed as not protected by copyright in
-the U.S. unless a copyright notice is included. Thus, we do not
-necessarily keep eBooks in compliance with any particular paper
-edition.
-
-Most people start at our website which has the main PG search
-facility: www.gutenberg.org
-
-This website includes information about Project Gutenberg-tm,
-including how to make donations to the Project Gutenberg Literary
-Archive Foundation, how to help produce our new eBooks, and how to
-subscribe to our email newsletter to hear about new eBooks.
diff --git a/old/66286-0.zip b/old/66286-0.zip
deleted file mode 100644
index 4b1c103..0000000
--- a/old/66286-0.zip
+++ /dev/null
Binary files differ
diff --git a/old/66286-h.zip b/old/66286-h.zip
deleted file mode 100644
index 6809311..0000000
--- a/old/66286-h.zip
+++ /dev/null
Binary files differ
diff --git a/old/66286-h/66286-h.htm b/old/66286-h/66286-h.htm
deleted file mode 100644
index 2e6d81e..0000000
--- a/old/66286-h/66286-h.htm
+++ /dev/null
@@ -1,4724 +0,0 @@
-<!DOCTYPE html PUBLIC "-//W3C//DTD XHTML 1.0 Strict//EN"
- "http://www.w3.org/TR/xhtml1/DTD/xhtml1-strict.dtd">
-<html xmlns="http://www.w3.org/1999/xhtml" xml:lang="en" lang="en">
- <head>
- <meta http-equiv="Content-Type" content="text/html;charset=utf-8" />
- <meta http-equiv="Content-Style-Type" content="text/css" />
- <title>
- Self-organizing Systems 1963, by James Emmett Garvey&mdash;A Project Gutenberg eBook
- </title>
- <link rel="coverpage" href="images/cover.jpg" />
- <style type="text/css">
-
-body { margin-left: 10%; margin-right: 10%; }
-
-h1,h2,h3,h4 { text-align: center; clear: both; }
-
-p { margin-top: .51em; text-align: justify; text-indent: 1.5em; margin-bottom: .49em; }
-p.no-indent { margin-top: .51em; text-align: justify; text-indent: 0em; margin-bottom: .49em;}
-p.author { margin-top: 1em; margin-right: 5%; text-align: right;}
-p.indent { text-indent: 1.5em;}
-p.neg-indent { text-indent: -1.5em; margin-left: 5%; margin-right: 5%; padding-left: 1.5em;}
-p.f90 { font-size: 90%; text-align: center; text-indent: 0em; }
-p.f120 { font-size: 120%; text-align: center; text-indent: 0em; }
-p.f150 { font-size: 150%; text-align: center; text-indent: 0em; }
-
-.fontsize_70 { font-size: 70%; }
-.fontsize_130 { font-size: 130%; }
-.fontsize_150 { font-size: 150%; }
-.fontsize_200 { font-size: 200%; }
-
-.space-above1 { margin-top: 1em; }
-.space-above2 { margin-top: 2em; }
-.space-above3 { margin-top: 3em; }
-.space-below1 { margin-bottom: 1em; }
-.space-below2 { margin-bottom: 2em; }
-.space-below3 { margin-bottom: 3em; }
-
-hr.chap {width: 65%; margin-left: 17.5%; margin-right: 17.5%; margin-top: 2em; margin-bottom: 2em;}
-hr.r5 {width: 5%; margin-top: 0.5em; margin-bottom: 0.5em;
- margin-left: 47.5%; margin-right: 47.5%;}
-hr.r25 {width: 25%; margin-top: 1em; margin-bottom: 1em;
- margin-left: 37.5%; margin-right: 37.5%;}
-
-div.chapter {page-break-before: always;}
-h2.nobreak {page-break-before: avoid;}
-
-ul.index { list-style-type: none; }
-li.isub1 {text-indent: 1em;}
-li.isub3 {text-indent: 3em;}
-li.isub4 {text-indent: 4em;}
-
-table { margin-left: auto; margin-right: auto; }
-.tdl {text-align: left;}
-.tdr {text-align: right;}
-.tdc {text-align: center;}
-.tdr_top {text-align: right; vertical-align: top;}
-.tdl_ws1 {text-align: left; vertical-align: top; padding-left: 1em;}
-.tdl_ws2 {text-align: left; vertical-align: top; padding-left: 2em;}
-
-.pagenum {
- position: absolute;
- left: 92%;
- font-size: smaller;
- text-align: right;
- font-style: normal;
- font-weight: normal;
- font-variant: normal;
-}
-
-.blockquot { margin-left: 10%; margin-right: 10%; }
-
-.bb {border-bottom: thin solid;}
-.bb2 {border-bottom: solid medium;}
-.bt2 {border-top: solid medium;}
-.br {border-right: thin solid;}
-.bbox {border: 2px solid;}
-
-.center {text-align: center;}
-.smcap {font-variant: small-caps;}
-
-img { max-width: 100%; height: auto; }
-
-.figcenter { margin: auto; text-align: center;
- page-break-inside: avoid; max-width: 100%; }
-
-div.figcontainer { clear: both; margin: 0em auto; text-align: center; max-width: 100%;}
-div.figsub { display: inline-block; margin: 1em 1em; vertical-align: top; max-width: 100%; text-align: center; }
-
-.footnotes {border: 1px dashed;}
-.footnote {margin-left: 10%; margin-right: 10%; font-size: 0.9em;}
-.footnote .label {position: absolute; right: 84%; text-align: right;}
-.fnanchor {
- vertical-align: super;
- font-size: .8em;
- text-decoration:
- none;
-}
-
-.transnote {background-color: #E6E6FA;
- color: black;
- font-size:smaller;
- padding:0.5em;
- margin-bottom:5em;
- font-family:sans-serif, serif; }
-
-.ws2 {display: inline; margin-left: 0em; padding-left: 2em;}
-.ws3 {display: inline; margin-left: 0em; padding-left: 3em;}
-.ws4 {display: inline; margin-left: 0em; padding-left: 4em;}
-.ws6 {display: inline; margin-left: 0em; padding-left: 6em;}
-.ws7 {display: inline; margin-left: 0em; padding-left: 7em;}
-.ws9 {display: inline; margin-left: 0em; padding-left: 9em;}
-.ws10 {display: inline; margin-left: 0em; padding-left: 10em;}
-
- </style>
- </head>
-<body>
-
-<div style='text-align:center; font-size:1.2em; font-weight:bold'>The Project Gutenberg eBook of Self-Organizing Systems, 1963, by Various</div>
-
-<div style='display:block; margin:1em 0'>
-This eBook is for the use of anyone anywhere in the United States and
-most other parts of the world at no cost and with almost no restrictions
-whatsoever. You may copy it, give it away or re-use it under the terms
-of the Project Gutenberg License included with this eBook or online
-at <a href="https://www.gutenberg.org">www.gutenberg.org</a>. If you
-are not located in the United States, you will have to check the laws of the
-country where you are located before using this eBook.
-</div>
-
-<p style='display:block; margin-top:1em; margin-bottom:1em; margin-left:2em; text-indent:-2em'>Title: Self-Organizing Systems, 1963</p>
-
-<div style='display:block; margin-top:1em; margin-bottom:1em; margin-left:2em; text-indent:-2em'>Author: Various</div>
-
-<div style='display:block; margin-top:1em; margin-bottom:1em; margin-left:2em; text-indent:-2em'>Editor: James Emmett Garvey</div>
-
-<div style='display:block; margin:1em 0'>Release Date: September 13, 2021 [eBook #66286]</div>
-
-<div style='display:block; margin:1em 0'>Language: English</div>
-
-<div style='display:block; margin:1em 0'>Character set encoding: UTF-8</div>
-
-<div style='display:block; margin-left:2em; text-indent:-2em'>Produced by: Mark C. Orton and the Online Distributed Proofreading Team at https://www.pgdp.net</div>
-
-<div style='margin-top:2em; margin-bottom:4em'>*** START OF THE PROJECT GUTENBERG EBOOK SELF-ORGANIZING SYSTEMS, 1963 ***</div>
-
-<hr class="chap x-ebookmaker-drop" />
-<h1>SELF-ORGANIZING SYSTEMS<br />1963</h1>
-
-<p class="center space-above3"><b>Edited By</b></p>
-
-<p class="center"><b>JAMES EMMETT GARVEY</b></p>
-<p class="center space-below3"><i>Office of Naval Research<br />Pasadena, California</i></p>
-
-<p class="center space-below2"><b>ACR-96</b></p>
-
-<p class="center space-below2"><b>OFFICE OF NAVAL RESEARCH<br />
-DEPARTMENT OF THE NAVY<br />WASHINGTON, D.C.</b></p>
-
-<p class="f90">For sale by the Superintendent of Documents.<br />
-U.S. Government Printing Office<br />
-Washington, D.C., 20402—Price $1.50</p>
-
-<hr class="chap x-ebookmaker-drop" />
-
-<div class="chapter">
-<h2 class="nobreak">CONTENTS</h2>
-</div>
-
-<table border="0" cellspacing="0" summary="TOC" cellpadding="2" >
- <tbody><tr>
- <td class="tdl">Foreword</td>
- <td class="tdr"><a href="#Page_iv">iv</a></td>
- </tr><tr>
- <td class="tdl">The Ionic Hypothesis and Neuron Models</td>
- <td class="tdr"><a href="#Page_1">&nbsp;1</a></td>
- </tr><tr>
- <td class="tdl_ws1">—E. R. Lewis</td>
- <td class="tdr">&nbsp;</td>
- </tr><tr>
- <td class="tdl">Fields and Waves in Excitable Cellular Structures</td>
- <td class="tdr"><a href="#Page_19">19</a></td>
- </tr><tr>
- <td class="tdl_ws1">—R. M. Stewart</td>
- <td class="tdr">&nbsp;</td>
- </tr><tr>
- <td class="tdl">Multi-Layer Learning Networks</td>
- <td class="tdr"><a href="#Page_37">37</a></td>
- </tr><tr>
- <td class="tdl_ws1">—R. A. Stafford</td>
- <td class="tdr">&nbsp;</td>
- </tr><tr>
- <td class="tdl">Adaptive Detection of Unknown Binary Waveforms&emsp;&nbsp;</td>
- <td class="tdr"><a href="#Page_46">46</a></td>
- </tr><tr>
- <td class="tdl_ws1">—J. J. Spilker, Jr.</td>
- <td class="tdr">&nbsp;</td>
- </tr><tr>
- <td class="tdl">Conceptual Design of Self-Organizing Machines</td>
- <td class="tdr"><a href="#Page_52">52</a></td>
- </tr><tr>
- <td class="tdl_ws1">—P. A. Kleyn</td>
- <td class="tdr">&nbsp;</td>
- </tr><tr>
- <td class="tdl">A Topological Foundation for Self-Organization</td>
- <td class="tdr"><a href="#Page_65">65</a></td>
- </tr><tr>
- <td class="tdl_ws1">—R. I. Ścibor-Marchocki</td>
- <td class="tdr">&nbsp;</td>
- </tr><tr>
- <td class="tdl">On Functional Neuron Modeling</td>
- <td class="tdr"><a href="#Page_71">71</a></td>
- </tr><tr>
- <td class="tdl_ws1">—C. E. Hendrix</td>
- <td class="tdr">&nbsp;</td>
- </tr><tr>
- <td class="tdl">Selection of Parameters for Neural Net Simulations</td>
- <td class="tdr"><a href="#Page_76">76</a></td>
- </tr><tr>
- <td class="tdl_ws1">—R. K. Overton</td>
- <td class="tdr">&nbsp;</td>
- </tr><tr>
- <td class="tdl">Index of Invited Participants</td>
- <td class="tdr"><a href="#Page_77">77</a></td>
- </tr>
- </tbody>
-</table>
-
-<hr class="chap x-ebookmaker-drop" />
-
-<div class="chapter">
-<p><span class="pagenum" id="Page_iv">[Pg iv]</span></p>
-<h2 class="nobreak">FOREWORD</h2>
-</div>
-
-<p>The papers appearing in this volume were presented at a Symposium
-on Self-Organizing Systems, which was sponsored by the Office of
-Naval Research and held at the California Institute of Technology,
-Pasadena, California, on 14 November 1963. The Symposium was organized
-with the aim of providing a critical forum for the presentation and
-discussion of contemporary significant research efforts, with the
-emphasis on relatively uncommon approaches and methods in an early
-state of development. This aim and nature dictated that the Symposium
-be in effect a Working Group, with numerically limited invitational
-participation.</p>
-
-<p>The papers which were presented and discussed did in fact serve
-to introduce several relatively unknown approaches; some of the
-speakers were promising young scientists, others had become known for
-contributions in different fields and were as yet unrecognized for
-their recent work in self-organization. In addition, the papers as a
-collection provided a particularly broad, cross-disciplinary spectrum
-of investigations which possessed intrinsic value as a portrayal of
-the bases upon which this new discipline rests. Accordingly, it became
-obvious in retrospect that the information presented and discussed at
-the Symposium was of considerable interest—and should thus receive
-commensurate dissemination—to a much broader group of scientists and
-engineers than those who were able to participate directly in the
-meeting itself. This volume is the result of that observation; as an
-edited collection of the papers presented at the Symposium, it forms
-the Proceedings thereof. If it provides a useful reference for present
-and future investigators, as well as documenting the source of several
-new approaches, it will have fulfilled its intended purpose well.</p>
-
-<p>A Symposium which takes the nature of a Working Group depends for its
-utility especially upon effective commentary and critical analysis,
-and we commend all the participants for their contributions in this
-regard. It is appropriate, further, to acknowledge the contributions
-to the success of the Symposium made by the following: The California
-Institute of Technology for volunteering to act as host and for
-numerous supporting services; Professor Gilbert D. McCann, Director
-of the Willis Booth Computing Center at the California Institute of
-Technology, and the members of the technical and secretarial staffs of
-the Computing Center, who assumed the responsibility of acting as the
-immediate representatives of the Institute; the members of the Program
-Committee, who organized and led the separate sessions—Harold Hamilton
-<span class="pagenum" id="Page_v">[Pg v]</span>
-of General Precision, Joseph Hawkins of Ford Motor Company, Robert
-Stewart of Space-General, Peter Kleyn of Northrop, and Professor
-McCann; members of the Technical Information Division of the Naval
-Research Laboratory, who published these Proceedings; and especially
-the authors of the papers, which comprised the heart of the Symposium
-and subsequently formed this volume. To all of these the sponsors wish
-to express their very sincere appreciation.</p>
-
-<p class="author"><span class="smcap">James Emmett Garvey</span><span class="ws6">&nbsp;</span><br />
-<i>Office of Naval Research Branch Office</i><br />
-<i>Pasadena, California</i><span class="ws7">&nbsp;</span></p>
-
-<p class="author"><span class="smcap">Margo A. Sass</span><span class="ws10">&nbsp;</span><br />
-<i>Office of Naval Research</i><span class="ws6">&nbsp;</span><br />
-<i>Washington, D.C.</i><span class="ws9">&nbsp;</span></p>
-
-<hr class="chap x-ebookmaker-drop" />
-
-<div class="chapter">
-<span class="pagenum" id="Page_1">[Pg 1]</span>
-<h2 class="nobreak">The Ionic Hypothesis and Neuron Models</h2>
-</div>
-
-<p class="f120"><b><span class="smcap">E. R. Lewis</span></b></p>
-
-<p class="center space-below1"><i>Librascope Group, General Precision, Inc.<br />
-Research and Systems Center<br />Glendale, California</i></p>
-
-<div class="blockquot">
-<p>The measurements of Hodgkin and Huxley were aimed at revealing the
-mechanism of generation and propagation of the all-or-none spike. Their
-results led to the Modern Ionic Hypothesis. Since the publication of
-their papers in 1952, advanced techniques with microelectrodes have led
-to the discovery of many modes of subthreshold activity not only in the
-axon but also in the somata and dendrites of neurons. This activity
-includes synaptic potentials, local response potentials, and pacemaker
-potentials.</p>
-
-<p>We considered the question, “Can this activity also be explained
-in terms of the Hodgkin-Huxley Model?” To seek an answer, we have
-constructed an electronic analog based on the ionic hypothesis and
-designed around the data of Hodgkin and Huxley. Synaptic inputs were
-simulated by simple first-order or second-order networks connected
-directly to simulated conductances (potassium or sodium). The analog
-has, with slight parameter adjustments, produced all modes of threshold
-and subthreshold activity.</p>
-</div>
-
-<h3>INTRODUCTION</h3>
-
-<p>In recent years physiologists have become quite adept at probing
-into neurons with intracellular microelectrodes. They are now able,
-in fact, to measure (a) the voltage change across the postsynaptic
-membrane elicited by a single presynaptic impulse (see, for examples,
-references <a href="#REF_A_1">1</a> and <a href="#REF_A_2">2</a>)
-and (b) the voltage-current characteristics across a
-localized region of the nerve cell membrane <a href="#REF_A_3">(3)</a>,
-<a href="#REF_A_4">(4)</a>, <a href="#REF_A_5">(5)</a>, <a href="#REF_A_6">(6)</a>.
-With microelectrodes, physiologists have been able to examine
-not only the all-or-none spike generating and propagating properties
-of axons but also the electrical properties of somatic and dendritic
-structures in individual neurons. The resulting observations have
-led many physiologists to believe that the individual nerve cell is
-a potentially complex information-processing system far removed from
-the simple two-state device envisioned by many early modelers. This
-new concept of the neuron is well summarized by Bullock in his 1959
-<span class="pagenum" id="Page_2">[Pg 2]</span>
-<i>Science</i> article <a href="#REF_A_10">(10)</a>. In the light of recent physiological
-literature, one cannot justifiably omit the diverse forms of somatic
-and dendritic behavior when assessing the information-processing
-capabilities of single neurons. This is true regardless of the
-means of assessment—whether one uses mathematical idealizations,
-electrochemical models, or electronic analogs. We have been interested
-specifically in electronic analogs of the neuron; and in view of the
-widely diversified behavior which we must simulate, our first goal has
-been to find a unifying concept about which to design our analogs. We
-believe we have found such a concept in the Modern Ionic Hypothesis,
-and in this paper we will discuss an electronic analog of the neuron
-which was based on this hypothesis and which simulated not only the
-properties of the axon but also the various subthreshold properties of
-the somata and dendrites of neurons.</p>
-
-<p>We begin with a brief summary of the various types of subthreshold
-activity which have been observed in the somatic and dendritic
-structures of neurons. This is followed by a brief discussion of the
-Hodgkin-Huxley data and of the Modern Ionic Hypothesis. An electronic
-analog based on the Hodgkin-Huxley data is then introduced, and we show
-how this analog can be used to provide all of the various types of
-somatic and dendritic activity.</p>
-
-<h3>SUBTHRESHOLD ELECTRICAL ACTIVITY<br /> IN NEURONS</h3>
-
-<p>In studying the recent literature in neurophysiology, one is
-immediately struck by the diversity in form of both elicited and
-spontaneous electrical activity in the single nerve cell. This applies
-not only to the temporal patterns of all-or-none action potentials
-but also to the graded somatic and dendritic potentials. The synaptic
-membrane of a neuron, for example, is often found to be electrically
-inexcitable and thus incapable of producing an action potential; yet
-the graded, synaptically induced potentials show an amazing diversity
-in form. In response to a presynaptic impulse, the postsynaptic
-membrane may become hyperpolarized (inhibitory postsynaptic potential),
-depolarized (excitatory postsynaptic potential), or remain at the
-resting potential but with an increased permeability to certain ions
-(a form of inhibition). The form of the postsynaptic potential in
-response to an isolated presynaptic spike may vary from synapse to
-synapse in several ways, as shown in <a href="#FIG_1A">Figure 1</a>. Following a
-presynaptic spike, the postsynaptic potential typically rises with some delay to
-a peak value and then falls back toward the equilibrium or resting
-potential. Three potentially important factors are the delay time
-(synaptic delay), the peak amplitude (spatial weighting of synapse),
-and the rate of fall toward the equilibrium potential (temporal
-weighting of synapse). The responses of a synapse to individual spikes
-in a volley may be progressively enhanced (facilitation), diminished
-(antifacilitation), or neither <a href="#REF_A_1">(1)</a>, <a href="#REF_A_2">(2)</a>,
-<a href="#REF_A_7">(7)</a>, <a href="#REF_A_8">(8)</a>. Facilitation may be
-in the form of progressively increased peak amplitude, or in the form
-of progressively decreased rate of fall (<a href="#FIG_2A">see Figure 2</a>). The
-time course and magnitude of facilitation or antifacilitation may very well be
-important synaptic parameters. In addition, the postsynaptic membrane
-sometimes exhibits excitatory or inhibitory aftereffects (or both) on
-cessation of a volley of presynaptic spikes <a href="#REF_A_2">(2)</a>,
-<a href="#REF_A_7">(7)</a>; and the time
-course and magnitude of the aftereffects may be important parameters.
-Clearly, even if one considers the synaptic potentials alone, he is
-faced with an impressive variety of responses. Examples of the various
-types of postsynaptic responses may be found in the literature, but for
-purposes of the present discussion the idealized wave forms in <a href="#FIG_2A">Figure 2</a>
-will demonstrate the diversity of electrical behavior with which one is faced.
-<span class="pagenum" id="Page_3">[Pg 3]</span></p>
-
-<div class="figcenter">
- <img id="FIG_1A" src="images/i_010a.jpg" alt="" width="600" height="252" />
- <img src="images/i_010b.jpg" alt="" width="600" height="168" />
- <img src="images/i_010c.jpg" alt="" width="600" height="88" />
- <img src="images/i_010d.jpg" alt="" width="600" height="135" />
- <p class="f120 space-below2">Figure 1—Excitatory postsynaptic potentials
- in response to a single presynaptic spike</p>
-</div>
-<hr class="r25 x-ebookmaker-drop" />
-<p><span class="pagenum" id="Page_4">[Pg 4]</span></p>
-<div class="figcenter">
- <img id="FIG_2A" src="images/i_011a.jpg" alt="" width="600" height="91" />
- <img src="images/i_011b.jpg" alt="" width="600" height="110" />
- <img src="images/i_011c.jpg" alt="" width="600" height="118" />
- <img src="images/i_011d.jpg" alt="" width="600" height="184" />
- <img src="images/i_011e.jpg" alt="" width="600" height="109" />
- <p class="f120 space-below2">Figure 2—Idealized postsynaptic potentials</p>
-</div>
-
-<p><span class="pagenum" id="Page_5">[Pg 5]</span>
-In addition to synaptically induced potentials, low-frequency,
-spontaneous potential fluctuations have been observed in many neurons
-<a href="#REF_A_2">(2)</a>, <a href="#REF_A_7">(7)</a>, <a href="#REF_A_9">(9)</a>,
-<a href="#REF_A_10">(10)</a>, <a href="#REF_A_11">(11)</a>. These fluctuations,
-generally referred to as pacemaker potentials, are usually rhythmic and may be
-undulatory or more nearly saw-toothed in form. The depolarizing phase may be
-accompanied by a spike, a volley of spikes, or no spikes at all.
-Pacemaker frequencies have been noted from ten or more cycles per
-second down to one cycle every ten seconds or more. Some idealized
-pacemaker wave forms are shown in <a href="#FIG_3A">Figure 3</a>.</p>
-
-<div class="figcenter">
- <img id="FIG_3A" src="images/i_012a.jpg" alt="" width="600" height="118" />
- <img src="images/i_012b.jpg" alt="" width="600" height="158" />
- <img src="images/i_012c.jpg" alt="" width="600" height="132" />
- <p class="f120 space-below2">Figure 3—Idealized pacemaker potentials</p>
-</div>
-<hr class="r25 x-ebookmaker-drop" />
-<p><span class="pagenum" id="Page_6">[Pg 6]</span></p>
-<div class="figcenter">
- <img id="FIG_4A" src="images/i_013a.jpg" alt="" width="600" height="226" />
- <img src="images/i_013b.jpg" alt="" width="600" height="331" />
- <p class="f120 space-below2">Figure 4—Graded response</p>
-</div>
-
-<p>Bullock <a href="#REF_A_7">(7)</a>, <a href="#REF_A_10">(10)</a>,
-<a href="#REF_A_12">(12)</a>, <a href="#REF_A_13">(13)</a> has demonstrated the
-existence of a third type of subthreshold response, which he calls the graded
-response. While the postsynaptic membrane is quite often electrically
-inexcitable, other regions of the somatic and dendritic membranes
-appear to be moderately excitable. It is in these regions that Bullock
-observes the graded response. If one applies a series of pulsed voltage
-stimuli to the graded-response region, the observed responses would be
-similar to those shown in <a href="#FIG_4A">Figure 4A</a>. Plotting the peak response
-voltage as a function of the stimulus voltage would result in a curve similar
-to that in <a href="#FIG_4A">Figure 4B</a> (<a href="#REF_A_3">see Ref. 3, page 4</a>).
-<span class="pagenum" id="Page_7">[Pg 7]</span>
-For small values of input voltage, the response curve is linear; the membrane is passive.
-As the stimulus voltage is increased, however, the response becomes more and
-more disproportionate. The membrane is actively amplifying the stimulus
-potential. At even higher values of stimulus potential, the system
-becomes regenerative; and a full action potential results. The peak
-amplitude of the response depends on the duration of the stimulus as
-well as on the amplitude. It also depends on the rate of application of
-the stimulus voltage. If the stimulus potential is a voltage ramp, for
-example, the response will depend on the slope of the ramp. If the rate
-of rise is sufficiently low, the membrane will respond in a passive
-manner to voltages much greater than the spike threshold for suddenly
-applied voltages. In other words, the graded-response regions appear to
-accommodate to slowly varying potentials.</p>
-
-<p>In terms of functional operation, we can think of the synapse as a
-transducer. The input to this transducer is a spike or series of spikes
-in the presynaptic axon. The output is an accumulative, long-lasting
-potential which in some way (perhaps not uniquely) represents the
-pattern of presynaptic spikes. The pacemaker appears to perform the
-function of a clock, producing periodic spikes or spike bursts or
-producing periodic changes in the over-all excitability of the neuron.
-The graded-response regions appear to act as nonlinear amplifiers and,
-occasionally, spike initiators. The net result of this electrical
-activity is transformed into a series of spikes which originate at
-spike initiation sites and are propagated along axons to other neurons.
-The electrical activity in the neuron described above is summarized in
-the following outline (taken in part from <a href="#REF_A_7">Bullock (7)</a>):</p>
-
-<ul class="index">
-<li class="isub1">1. Synaptic Potentials</li>
-<li class="isub3">a. Excitatory or inhibitory</li>
-<li class="isub3">b. Facilitated, antifacilitated, or neither</li>
-<li class="isub3">c. With excitatory aftereffect, inhibitory</li>
-<li class="isub4">aftereffect, neither, or both</li>
-<li class="isub1">2. Pacemaker Potentials</li>
-<li class="isub3">a. Relaxation type, undulatory type,</li>
-<li class="isub4">or none at all</li>
-<li class="isub3">b. Producing single spike, spike burst,</li>
-<li class="isub4">or no spikes</li>
-<li class="isub3">c. Rhythmic or sporadic</li>
-<li class="isub1">3. Graded Response (rate sensitive)</li>
-<li class="isub1">4. Spike Initiation</li>
-</ul>
-<p class="space-below2"><span class="pagenum" id="Page_8">[Pg 8]</span></p>
-
-<h3>THE MODERN IONIC HYPOTHESIS</h3>
-
-<p>Hodgkin, Huxley, and Katz <a href="#REF_A_3">(3)</a> and Hodgkin and Huxley
-<a href="#REF_A_14">(14)</a>, <a href="#REF_A_15">(15)</a>, <a href="#REF_A_16">(16)</a>,
-in 1952, published a series of papers describing detailed measurements
-of voltage, current, and time relationships in the giant axon of
-the squid (<i>Loligo</i>). Hodgkin and Huxley <a href="#REF_A_17">(17)</a>
-consolidated and formalized these data into a set of simultaneous differential
-equations describing the hypothetical time course of events during
-spike generation and propagation. The hypothetical system which these
-equations describe is the basis of the Modern Ionic Hypothesis.</p>
-
-<p>The system proposed by Hodgkin and Huxley is basically one of dynamic
-opposition of ionic fluxes across the axon membrane. The membrane
-itself forms the boundary between two liquid phases—the intracellular
-fluid and the extracellular fluid. The intracellular fluid is rich in
-potassium ions and immobile organic anions, while the extracellular
-fluid contains an abundance of sodium ions and chloride ions. The
-membrane is slightly permeable to the potassium, sodium, and chloride
-ions; so these ions tend to diffuse across the membrane. When the
-axon is inactive (not propagating a spike), the membrane is much more
-permeable to chloride and potassium ions than it is to sodium ions.
-In this state, in fact, sodium ions are actively transported from the
-inside of the membrane to the outside at a rate just sufficient to
-balance the inward leakage. The relative sodium ion concentrations on
-both sides of the membrane are thus fixed by the active transport rate,
-and the net sodium flux across the membrane is effectively zero. The
-potassium ions, on the other hand, tend to move out of the cell; while
-chloride ions tend to move into it. The inside of the cell thus becomes
-negative with respect to the outside. When the potential across the
-membrane is sufficient to balance the inward diffusion of chloride with
-an equal outward drift, and the outward diffusion of potassium with an
-inward drift (and possibly an inward active exchange), equilibrium is
-established. The equilibrium potential is normally in the range of 60
-to 65 millivolts.</p>
-
-<p>The resting neural membrane is thus polarized, with the inside
-approximately 60 millivolts negative with respect to the outside.
-Most of the Hodgkin-Huxley data is based on measurements of the
-transmembrane current in response to an imposed stepwise reduction
-(depolarization) of membrane potential. By varying the external
-ion concentrations, Hodgkin and Huxley were able to resolve the
-transmembrane current into two “active” components, the potassium ion
-current and the sodium ion current. They found that while the membrane
-<span class="pagenum" id="Page_9">[Pg 9]</span>
-permeabilities to chloride and most other inorganic ions were
-relatively constant, the permeabilities to both potassium and sodium
-were strongly dependent on membrane potential. In response to a
-suddenly applied (step) depolarization, the sodium permeability rises
-rapidly to a peak and then declines exponentially to a steady value.
-The potassium permeability, on the other hand, rises with considerable
-delay to a value which is maintained as long as the membrane remains
-depolarized. The magnitudes of both the potassium and the sodium
-permeabilities increase monotonically with increasing depolarization.
-A small imposed depolarization will result in an immediately
-increased sodium permeability. The resulting increased influx of
-sodium ions results in further depolarization; and the process
-becomes regenerative, producing the all-or-none action potential.
-At the peak of the action potential, the sodium conductance begins
-to decline, while the delayed potassium conductance is increasing.
-Recovery is brought about by an efflux of potassium ions, and both
-ionic permeabilities fall rapidly as the membrane is repolarized.
-The potassium permeability, however, falls less rapidly than that of
-sodium. This is basically the explanation of the all-or-none spike
-according to the Modern Ionic Hypothesis.</p>
-
-<div class="figcenter">
- <img id="FIG_5A" src="images/i_016.jpg" alt="" width="600" height="365" />
- <p class="f120 space-below2">Figure 5—Hodgkin-Huxley representation
- of small area of axon membrane</p>
-</div>
-<div class="figcenter">
- <img id="FIG_6A" src="images/i_017.jpg" alt="" width="600" height="540" />
- <p class="f120 space-below2">Figure 6—Typical responses of sodium conductance and
-potassium conductance to imposed step depolarization</p>
-</div>
-
-<p>By defining the net driving force on any given ion species as the
-difference between the membrane potential and the equilibrium potential
-for that ion and describing permeability changes in terms of equivalent
-electrical conductance changes, Hodgkin and Huxley reduced the ionic
-<span class="pagenum" id="Page_10">[Pg 10]</span>
-model to the electrical equivalent in <a href="#FIG_5A">Figure 5</a>. The important
-dynamic variables in this equivalent network are the sodium conductance
-<big>(G{Na})</big> and the potassium conductance <big>(G{K})</big>. The change in
-the sodium conductance in response to a step depolarization is shown in <a href="#FIG_6A">Figure 6B</a>.
-This change can be characterized by seven voltage dependent parameters:
-<span class="pagenum" id="Page_11">[Pg 11]</span></p>
-
-<div class="blockquot">
-<p class="neg-indent">1. Delay time—generally much less than 1 msec</p>
-
-<p class="neg-indent">2. Rise time—1 msec or less</p>
-
-<p class="neg-indent">3. Magnitude of peak conductance—increases
-monotonically with increasing depolarization</p>
-
-<p class="neg-indent">4. Inactivation time constant—decreases
-monotonically with increasing depolarization.</p>
-
-<p class="neg-indent">5. Time constant of recovery from
-inactivation—incomplete data</p>
-
-<p class="neg-indent">6. Magnitude of steady-state
-conductance—increases monotonically with increasing depolarization</p>
-
-<p class="neg-indent">7. Fall time on sudden repolarization—less than 1
-msec.</p>
-</div>
-
-<p><a href="#FIG_6A">Figure 6B</a> shows the potassium conductance change in
-response to an imposed step depolarization. Four parameters are sufficient to
-characterize this response:</p>
-
-<div class="blockquot">
-<p class="neg-indent">1. Delay time—decreases monotonically with
-increasing depolarization</p>
-
-<p class="neg-indent">2. Rise time—decreases monotonically with
-increasing depolarization</p>
-
-<p class="neg-indent">3. Magnitude of steady-state
-conductance—increases monotonically with increasing depolarization</p>
-
-<p class="neg-indent">4. Fall time on sudden repolarization—8 msec or
-more, decreases slightly with increasing depolarization.</p>
-</div>
-
-<p>In addition to the aforementioned parameters, the transient portion of
-the sodium conductance appears to exhibit an accommodation to slowly
-varying membrane potentials. The time constants of accommodation appear
-to be those of inactivation or recovery from inactivation—depending on
-the direction of change in the membrane potential <a href="#REF_A_18">(18)</a>.
-The remaining elements in the Hodgkin-Huxley model are constant and are listed below:</p>
-
-<ul class="index">
-<li class="isub1">1. Potassium potential—80 to 85 mv (inside negative)</li>
-<li class="isub1">2. Sodium potential—45 to 50 mv (inside positive)</li>
-<li class="isub1">3. Leakage potential—38 to 43 mv (inside negative)</li>
-<li class="isub1">4. Leakage conductance—approx. 0.23 millimhos/cm²</li>
-<li class="isub1">5. Membrane capacitance—approx. 1 μf/cm²</li>
-<li class="isub1">6. Resting potential—60 to 65 mv</li>
-<li class="isub1">7. Spike amplitude—approx. 100 mv</li>
-</ul>
-
-<h3 class="space-above2">ELECTRONIC SIMULATION OF THE HODGKIN-HUXLEY MODEL</h3>
-
-<div class="figcenter">
- <img id="FIG_7A" src="images/i_019.jpg" alt="" width="600" height="561" />
- <p class="f120 space-below2">Figure 7—System diagram for electronic
- simulation of the Hodgkin-Huxley model</p>
-</div>
-
-<p>Given a suitable means of generating the conductance functions,
-<big>G<sub><i>Na</i></sub>(v,t)</big> and <big>G<sub><i>K</i></sub>(v,t)</big>, one can
-readily stimulate the essential aspects of the Modern Ionic Hypothesis. If we wish to
-do this electronically, we have two problems. First, we must synthesize a
-network whose input is the membrane potential and whose output is a
-<span class="pagenum" id="Page_12">[Pg 12]</span>
-voltage or current proportional to the desired conductance function.
-Second, we must transform the output from a voltage or current to
-an effective electronic conductance. The former implies the need
-for nonlinear, active filters, while the latter implies the need
-for multipliers. The basic block diagram is shown in <a href="#FIG_7A">Figure 7</a>.
-Several distinct realizations of this system have been developed in
-our laboratory, and in each case the results were the same. With
-parameters adjusted to closely match the data of Hodgkin and Huxley,
-the electronic model exhibits all of the important properties of the
-axon. It produces spikes of 1 to 2 msec duration with a threshold of
-approximately 5% to 10% of the spike amplitude. The applied stimulus is
-<span class="pagenum" id="Page_13">[Pg 13]</span>
-generally followed by a prepotential, then an active rise of less than
-1 msec, followed by an active recovery. The after-depolarization
-generally lasts several msec, followed by a prolonged
-after-hyperpolarization. The model exhibits the typical
-strength-duration curve, with rheobase of 5% to 10% of the spike
-amplitude. For sufficiently prolonged sodium inactivation (long time
-constant of recovery from inactivation), the model also exhibits an
-effect identical to classical Wedensky inhibition <a href="#REF_A_18">(18)</a>.
-Thus, as would be expected, the electronic model simulates very well the
-electrical properties of the axon.</p>
-
-<p>In addition to the axon properties, however, the electronic model is
-able to reproduce all of the somatic and dendritic activity outlined
-in the section on subthreshold activity. Simulation of the pacemaker
-and graded-response potentials is accomplished without additional
-circuitry. In the case of synaptically induced potentials, however,
-auxiliary networks are required. These networks provide additive terms
-to the variable conductances in accordance with current notions on
-synaptic transmission <a href="#REF_A_19">(19)</a>. Two types of networks have
-been used. In both, the inputs are simulated presynaptic spikes, and in both the
-outputs are the resulting simulated chemical transmitter concentration.
-In both, the transmitter substance was assumed to be injected at a
-constant rate during a presynaptic spike and subsequently inactivated
-in the presence of an enzyme. One network simulates a first-order
-chemical reaction, where the enzyme concentration is effectively
-constant. The other simulates a second-order chemical reaction,
-where the enzyme concentration is assumed to be reduced during the
-inactivation process. For simulation of an excitatory synapse, the
-output of the auxiliary network is added directly to <big>G<sub><i>Na</i></sub></big> in the
-electronic model. For inhibition, it is added to <big>G<sub><i>K</i></sub></big>. With the
-parameters of the electronic membrane model set at the values measured
-by Hodgkin and Huxley, we have attempted to simulate synaptic activity
-with the aid of the two types of auxiliary networks. In the case of
-the simulated first-order reaction, the excitatory synapse exhibits
-facilitation, antifacilitation, or neither—depending on the setting
-of a single parameter, the transmitter inactivation rate (<i>i.e.</i>,
-the effective enzyme concentration). This parameter would appear,
-in passing, to be one of the most probable synaptic variables. In
-this case, the mechanisms for facilitation and antifacilitation are
-contained in the simulated postsynaptic membrane. Facilitation is due
-to the nonlinear dependence of <big>G<sub><i>Na</i></sub></big> on membrane potential, while
-antifacilitation is due to inactivation of <big>G<sub><i>Na</i></sub></big>. The occurrence
-of one form of response or the other is determined by the relative
-importance of the two mechanisms <a href="#REF_A_18">(18)</a>.
-Grundfest <a href="#REF_A_20">(20)</a> has mentioned
-<span class="pagenum" id="Page_14">[Pg 14]</span>
-both of these mechanisms as potentially facilitory and antifacilitory,
-respectively. The simulated inhibitory synapse with the first order
-input is capable of facilitation <a href="#REF_A_18">(18)</a>, but no antifacilitation
-has been observed. Again, the presence or absence of facilitation is determined
-by the inactivation rate.</p>
-
-<p>With the simulated second-order reaction, both excitatory and
-inhibitory synapses exhibit facilitation. In this case, two facilitory
-mechanisms are present—one in the postsynaptic membrane and one
-in the nonconstant transmitter inactivation reaction. The active
-membrane currents can, in fact, be removed; and this system will still
-exhibit facilitation. With the second-order auxiliary network, the
-presence of excitatory facilitation, antifacilitation, or neither
-depends on the initial, or resting, transmitter inactivation rate. The
-synaptic behavior also depends parametrically on the simulated enzyme
-reactivation rate. Inhibitory antifacilitation can be introduced with
-either type of auxiliary network by limiting the simulated presynaptic
-transmitter supply.</p>
-
-<p>Certain classes of aftereffects are inherent in the mechanisms of the
-Ionic Hypothesis. In the electronic model, aftereffects are observed
-following presynaptic volleys with either type of auxiliary network.
-Following a volley of spikes into the simulated excitatory synapse,
-for example, rebound hyperpolarization may or may not occur depending
-on the simulated transmitter inactivation rate. If the inactivation
-rate is sufficiently high, rebound will occur. This rebound can be
-monophasic (inhibitory phase only) or polyphasic (successive cycles
-of excitation and inhibition). Following a volley of spikes into the
-simulated inhibitory synapse, rebound depolarization may or may not
-occur depending on the simulated transmitter inactivation rate. This
-rebound can also be monophasic or polyphasic. Sustained postexcitatory
-depolarization and sustained postinhibitory hyperpolarization <a href="#REF_A_2">(2)</a>
-have been achieved in the model by making the transmitter inactivation rate
-sufficiently low.</p>
-
-<p>The general forms of the postsynaptic potentials simulated with
-the electronic model are strikingly similar to those published in
-the literature for real neurons. The first-order auxiliary network
-produces facilitation of a form almost identical to that shown by Otani
-and Bullock <a href="#REF_A_8">(8)</a> while the second-order auxiliary network produces
-facilitation of the type shown by Chalazonitis and Arvanitake <a href="#REF_A_2">(2)</a>.
-The excitatory antifacilitation is almost identical to that shown by
-Hagiwara and Bullock <a href="#REF_A_1">(1)</a> in both form and dependence on presynaptic
-spike frequency. In every case, the synaptic behavior is determined by
-the effective rate of transmitter inactivation, which in real neurons
-<span class="pagenum" id="Page_15">[Pg 15]</span>
-would presumably be directly proportional to the effective
-concentration of inactivating enzyme at the synapse.</p>
-
-<p>Pacemaker potentials are easily simulated with the electronic model
-without the use of auxiliary networks. This is achieved either by
-inserting a large, variable shunt resistor across the simulated
-membrane (<a href="#FIG_5A">see Figure 5</a>) or by allowing a small sodium current
-leakage at the resting potential. With the remaining parameters of the
-model set as close as possible to the values determined by Hodgkin
-and Huxley, the leakage current induces low-frequency, spontaneous
-spiking. The spike frequency increases monotonically with increasing
-leakage current. In addition, if the sodium conductance inactivation
-is allowed to accumulate over several spikes, periodic spike pairs
-and spike bursts will result. Subthreshold pacemaker potentials have
-also been observed in the model, but with parameter values set close
-to the Hodgkin-Huxley data these are generally higher in frequency
-than pacemaker potentials in real neurons. It is interesting that
-a pacemaker mode may exist in the absence of the simulated sodium
-conductance. It is a very high-frequency mode (50 cps or more)
-and results from the alternating dominance of potassium current
-and chloride (or leakage ion) current in determining the membrane
-potential. The significance of this mode cannot be assessed until
-better data is available for the potassium conductance at low levels
-of depolarization in real neurons. In general, as far as the model is
-concerned, pacemaker potentials are possible because the potassium
-conductance is delayed in both its rise with depolarization and its
-fall with repolarization.</p>
-
-<p>Rate sensitive graded response has also been observed in the electronic
-model. The rate sensitivity—or accommodation—is due to the sodium
-conductance inactivation. The response of the model to an imposed ramp
-depolarization was discussed in <a href="#REF_A_18">Reference 18</a>. At this time,
-several alternative model parameters could be altered to bring about reduced
-electrical excitability. None of the parameter changes was very
-satisfying, however, because none of them was in any way justified by
-physiological data. We have since found that the membrane capacitance,
-a plausible parameter in view of recent physiological findings, can
-completely determine the electrical excitability. Thus, with the
-capacitance determined by Hodgkin and Huxley (1 microfarad per cm²),
-the model exhibits excitability characteristic of the axon. As the
-capacitance is increased, the model becomes less excitable until, with
-10 or 12 μμf, it is effectively inexcitable. Thus, with an increased
-<span class="pagenum" id="Page_16">[Pg 16]</span>
-capacitance—but with all the remaining parameters set as close as
-possible to the Hodgkin-Huxley values—the electronic model exhibits
-the characteristics of Bullock’s graded-response regions.</p>
-
-<p>Whether membrane capacitance is the determining factor in real neurons
-is, of course, a matter of speculation. Quite a controversy is raging
-over membrane capacity measurements (<a href="#REF_A_21">see Rall (21)</a>), but the
-evidence indicates that the capacity in the soma is considerably greater than
-that in the axon <a href="#REF_A_6">(6)</a>, <a href="#REF_A_22">(22)</a>.</p>
-
-<p>It should be added that increasing the capacitance until the membrane
-model becomes inexcitable has little effect on the variety of available
-simulated synaptic responses. Facilitation, antifacilitation, and
-rebound are still present and still depend on the transmitter
-inactivation rate. Thus, in the model, we can have a truly inexcitable
-membrane which nevertheless utilizes the active membrane conductances
-to provide facilitation or antifacilitation, and rebound. The simulated
-subthreshold pacemaker potentials are much more realistic with the
-increased capacitance, being lower in frequency and more natural in form.</p>
-
-<p>In one case, the electronic model predicted behavior which was
-subsequently reported in real neurons. This was in respect to the
-interaction of synaptic potentials and pacemaker potential. It was
-noted in early experiments that when the model was set in a pacemaker
-mode, and periodic spikes were applied to the simulated inhibitory
-synapse, the pacemaker frequency could be modified; and, in fact,
-it would tend to lock on to the stimulus frequency. This produced
-a paradoxical effect whereby the frequency of spontaneous spikes
-was actually increased by increasing the frequency of inhibitory
-synaptic stimuli. At very low stimulus frequencies, the spontaneous
-pacemaker frequency was not appreciably perturbed. As the stimulus
-frequency was increased, and approached the basic pacemaker frequency,
-the latter tended to lock on and follow further increases in the
-stimulus frequency. When the stimulus frequency became too high for
-the pacemaker to follow, the latter decreased abruptly in frequency
-and locked on to the first subharmonic. As the stimulus frequency was
-further increased, the pacemaker frequency would increase, then skip
-to the next harmonic, then increase again, <i>etc.</i> This type of
-behavior was observed by Moore <i>et al.</i> <a href="#REF_A_23">(23)</a>
-in <i>Aplysia</i> and reported at the San Diego Symposium for Biomedical Electronics
-shortly after it was observed by the author in the electronic model.</p>
-
-<p>Thus, we have shown that an electronic analog with all parameters
-except membrane capacitance fixed at values close to those of Hodgkin
-and Huxley, can provide all of the normal threshold or axonal behavior
-<span class="pagenum" id="Page_17">[Pg 17]</span>
-and also all of the subthreshold somatic and dendritic behavior
-outlined on <a href="#Page_7">page 7</a>. Whether or not this is of physiological
-significance, it certainly provides a unifying basis for construction
-of electronic neural analogs. Simple circuits, based on the
-Hodgkin-Huxley model and providing all of the aforementioned behavior,
-have been constructed with ten or fewer inexpensive transistors with
-a normal complement of associated circuitry <a href="#REF_A_18">(18)</a>. In the
-near future we hope to utilize several models of this type to help assess the
-information-processing capabilities not only of individual neurons but
-also of small groups or networks of neurons.</p>
-
-<p class="f120 space-above1"><b>REFERENCES</b></p>
-<table border="0" cellspacing="0" summary="REFERENCES" cellpadding="2" >
- <tbody><tr>
- <td id="REF_A_1" class="tdr">1.</td>
- <td class="tdl_ws1">Hagiwara, S., and Bullock, T. H.</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">“Intracellular Potentials in Pacemaker
- and Integrative Neurons of the Lobster Cardiac Ganglion,”</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws2"><i>J. Cell and Comp. Physiol.</i> <b>50 (No. 1)</b>:25-48 (1957)</td>
- </tr><tr>
- <td id="REF_A_2" class="tdr">2.</td>
- <td class="tdl_ws1">Chalazonitis, N., and Arvanitaki, A.,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">“Slow Changes during and following Repetitive
- Synaptic Activation in Ganglion Nerve Cells,”</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws2"><i>Bull. Inst. Oceanogr. Monaco</i> <b>No. 1225</b>:1-23 (1961)</td>
- </tr><tr>
- <td id="REF_A_3" class="tdr">3.</td>
- <td class="tdl_ws1">Hodgkin, A. L., Huxley, A. F., and Katz, B.,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">“Measurement of Current-Voltage Relations in the
- Membrane of the Giant Axon of <i>Loligo</i>,”</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws2"><i>J. Physiol.</i> <b>116</b>:424-448 (1952)</td>
- </tr><tr>
- <td id="REF_A_4" class="tdr">4.</td>
- <td class="tdl_ws1">Hagiwara, S., and Saito, N.,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">“Voltage-Current Relations in Nerve Cell Membrane
- of Onchidium <i>verruculatum</i>,”</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws2"><i>J. Physiol.</i> <b>148</b>:161-179 (1959)</td>
- </tr><tr>
- <td id="REF_A_5" class="tdr">5.</td>
- <td class="tdl_ws1">Hagiwara, S., and Saito, N.,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">“Membrane Potential Change and Membrane Current in
- Supramedullary Nerve Cell of Puffer,”</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws2"><i>J. Neurophysiol.</i> <b>22</b>:204-221 (1959)</td>
- </tr><tr>
- <td id="REF_A_6" class="tdr">6.</td>
- <td class="tdl_ws1">Hagiwara, S.,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">“Current-Voltage Relations of Nerve Cell Membrane,”</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">“Electrical Activity of Single Cells,”</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws2">Igakushoin, Hongo, Tokyo (1960)</td>
- </tr><tr>
- <td id="REF_A_7" class="tdr">7.</td>
- <td class="tdl_ws1">Bullock, T. H.,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">“Parameters of Integrative Action of
- the Nervous System at the Neuronal Level,”</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws2"><i>Experimental Cell Research Suppl.</i> <b>5</b>:323-337 (1958)</td>
- </tr><tr>
- <td id="REF_A_8" class="tdr">8.</td>
- <td class="tdl_ws1">Otani, T., and Bullock, T. H.,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">“Effects of Presetting the Membrane Potential of the Soma
- of Spontaneous and Integrating Ganglion Cells,”</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws2"><i>Physiological Zoology</i> <b>32 (No. 2)</b>:104-114 (1959)</td>
- </tr><tr>
- <td id="REF_A_9" class="tdr">9.</td>
- <td class="tdl_ws1">Bullock, T. H., and Terzuolo, C. A.,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">“Diverse Forms of Activity in the Somata of
- Spontaneous and Integrating Ganglion Cells,”</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws2"><i>J. Physiol.</i> <b>138</b>:343-364 (1957)</td>
- </tr><tr>
- <td id="REF_A_10" class="tdr">10.</td>
- <td class="tdl_ws1">Bullock, T. H.,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">“Neuron Doctrine and Electrophysiology,”</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws2"><i>Science</i> <b>129 (No. 3355)</b>:997-1002 (1959)</td>
- </tr><tr>
- <td id="REF_A_11" class="tdr">11.</td>
- <td class="tdl_ws1">Chalazonitis, N., and Arvanitaki, A.,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">“Slow Waves and Associated Spiking in Nerve Cells of <i>Aplysia</i>,”</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws2"><i>Bull. Inst. Oceanogr. Monaco</i> <b>No. 1224</b>:1-15 (1961)</td>
- </tr><tr>
- <td id="REF_A_12" class="tdr">12.</td>
- <td class="tdl_ws1">Bullock, T. H.,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">“Properties of a Single Synapse in the Stellate Ganglion of Squid,”</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws2"><i>J. Neurophysiol.</i> <b>11</b>:343-364 (1948)</td>
- </tr><tr>
- <td id="REF_A_13" class="tdr">13.</td>
- <td class="tdl_ws1">Bullock, T. H.,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">“Neuronal Integrative Mechanisms,”</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">“Recent Advances in Invertebrate Physiology,”</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws2">Scheer, B. T., ed., Eugene, Oregon:Univ. Oregon Press 1957</td>
- </tr><tr>
- <td id="REF_A_14" class="tdr">14.</td>
- <td class="tdl_ws1">Hodgkin, A. L., and Huxley, A. F.,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">“Currents Carried by Sodium and Potassium Ions through
- the Membrane of the Giant Axon of Loligo,”</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws2"><i>J. Physiol.</i> <b>116</b>:449-472 (1952)</td>
- </tr><tr>
- <td id="REF_A_15" class="tdr">15.</td>
- <td class="tdl_ws1">Hodgkin, A. L., and Huxley, A. F.,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">“The Components of Membrane Conductance in
- the Giant Axon of <i>Loligo</i>,”</td>
- </tr><tr>
- <td class="tdr"><span class="pagenum" id="Page_18">[Pg 18]</span></td>
- <td class="tdl_ws2"><i>J. Physiol.</i> <b>116</b>:473-496 (1952)</td>
- </tr><tr>
- <td id="REF_A_16" class="tdr">16.</td>
- <td class="tdl_ws1">Hodgkin, A. L., and Huxley, A. F.,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">“The Dual Effect of Membrane Potential on Sodium Conductance
- in the Giant Axon of <i>Loligo</i>,”</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws2"><i>J. Physiol.</i> <b>116</b>:497-506 (1952)</td>
- </tr><tr>
- <td id="REF_A_17" class="tdr">17.</td>
- <td class="tdl_ws1">Hodgkin, A. L., and Huxley, A. F.,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">“A Quantitative Description of Membrane Current and its
- Application to Conduction and Excitation in Nerve,”</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws2"><i>J. Physiol.</i> <b>117</b>:500-544 (1952)</td>
- </tr><tr>
- <td id="REF_A_18" class="tdr">18.</td>
- <td class="tdl_ws1">Lewis, E. R.,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">“An Electronic Analog of the Neuron Based on the Dynamics of
- Potassium and Sodium Ion Fluxes,”</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">“Neural Theory and Modeling,”</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws2">R. F. Reiss, ed., Palo Alto, California:Stanford
- University Press, 1964</td>
- </tr><tr>
- <td id="REF_A_19" class="tdr">19.</td>
- <td class="tdl_ws1">Eccles, J. C.,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws2"><i>Physiology of Synapses</i>,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">Berlin:Springer-Verlag, 1963</td>
- </tr><tr>
- <td id="REF_A_20" class="tdr">20.</td>
- <td class="tdl_ws1">Grundfest, H.,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">“Excitation Triggers in Post-Junctional Cells,”</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">“Physiological Triggers,”</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws2">T. H. Bullock, ed., Washington, D.C.:American
- Physiological Society, 1955</td>
- </tr><tr>
- <td id="REF_A_21" class="tdr">21.</td>
- <td class="tdl_ws1">Rall, W.,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">“Membrane Potential Transients and Membrane Time
- Constants of Motoneurons,”</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws2"><i>Exp. Neurol.</i> <b>2</b>:503-532 (1960)</td>
- </tr><tr>
- <td id="REF_A_22" class="tdr">22.</td>
- <td class="tdl_ws1">Araki, T., and Otani, T.,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">“The Response of Single Motoneurones to Direct Stimulation,”</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws2"><i>J. Neurophysiol.</i> <b>18</b>:472-485 (1955)</td>
- </tr><tr>
- <td id="REF_A_23" class="tdr">23.</td>
- <td class="tdl_ws1">Moore, G. P., Perkel, D. H., and Segundo, J. P.,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">“Stability Patterns in Interneuronal Pacemaker Regulation,”</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws2"><i>Proceedings of the San Diego Symposium for Biomedical Engineering</i>,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">San Diego, California, 1963</td>
- </tr><tr>
- <td id="REF_A_24" class="tdr">24.</td>
- <td class="tdl_ws1">Eccles, J. C.,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1"><i>The Neurophysiological Basis of Mind</i>,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws2">Oxford:Clarendon Press, 1952</td>
- </tr>
- </tbody>
-</table>
-
-<hr class="chap x-ebookmaker-drop" />
-
-<div class="chapter">
-<p><span class="pagenum" id="Page_19">[Pg 19]</span></p>
-<h2 class="nobreak">Fields and Waves in Excitable<br />Cellular Structures</h2>
-</div>
-
-<p class="f120"><b><span class="smcap">R. M. STEWART</span></b></p>
-
-<p class="center space-below1"><i>Space General Corporation<br />El Monte, California</i></p>
-
-<div class="blockquot">
-<p>“Study of living processes by the physiological method only
-proceeded laboriously behind the study of non-living systems. Knowledge
-about respiration, for instance, began to become well organized
-as the study of combustion proceeded, since this is an analogous
-operation....”</p>
-
-<p class="author"><a href="#REF_B_24">J. Z. Young (24)</a></p>
-</div>
-
-<h3>INTRODUCTION</h3>
-
-<p>The study of electrical fields in densely-packed cellular media is
-prompted primarily by a desire to understand more fully the details
-of brain mechanism and its relation to behavior. Our work has
-specifically been directed toward an attempt to model such structures
-and mechanisms, using relatively simple inorganic materials.</p>
-
-<p>The prototype for such experiments is the “Lillie<a id="FNanchor_1" href="#Footnote_1" class="fnanchor">[1]</a>
-iron-wire nerve model.” Over a hundred years ago, it had been observed
-that visible waves were produced on the surface of a piece of iron
-submerged in nitric acid when and where the iron is touched by a piece
-of zinc. After a short period of apparent fatigue, the wire recovers
-and can again support a wave when stimulated. Major support for the
-idea that such impulses are in fact directly related to peripheral
-nerve impulses came from Lillie around 1920. Along an entirely
-different line, various persons have noted the morphological and
-dynamic similarity of dendrites in brain and those which sometimes
-grow by electrodeposition of metals from solution. Gordon Pask <a href="#REF_B_17">(17)</a>,
-especially, has pointed to this similarity and has discussed in a
-general way the concomitant possibility of a physical model for the
-persistent memory trace.</p>
-
-<p>By combining and extending such concepts and techniques, we hope to
-produce a macroscopic model of “gray matter,” the structural matrix of
-which will consist of a dense, homogeneously-mixed, conglomerate of
-<span class="pagenum" id="Page_20">[Pg 20]</span>
-small pellets, capable of supporting internal waves of excitation, of
-changing electrical behavior through internal fine-structure growth,
-and of forming temporal associations in response to peripheral shocks.</p>
-
-<p>A few experimenters have subsequently pursued the iron-wire
-nerve-impulse analogy further, hoping thereby to illuminate the
-mechanisms of nerve excitation, impulse transmission and recovery,
-but interest has generally been quite low. It has remained fairly
-undisturbed in the text books and lecture demonstrations of medical
-students, as a picturesque aid to their formal education. On the
-outer fringes of biology, still less interest has been displayed;
-the philosophical vitalists would surely be revolted by the idea of
-such models of mind and memory, and at the other end of the scale,
-contemporary computer engineers generally assume that a nerve cell
-operates much too slowly to be of any value. This lack of interest
-is certainly due, in part, to success in developing techniques of
-monitoring individual nerve fibers directly to the point that it is
-just about as easy to work with large nerve fibers (and even peripheral
-and spinal junctions) as it is to work with iron wires. Under such
-circumstances, the model has only limited value, perhaps just to the
-extent that it emphasizes the role of factors other than specific
-molecular structure and local chemical reactions in the dynamics of
-nerve action.</p>
-
-<p>When we leave the questions of impulse transmission on long fibers
-and peripheral junctions, however, and attempt to discuss the brain,
-there can be hardly any doubt that the development of a meaningful
-physical model technique would be of great value. Brain tissue is
-soft and sensitive, the cellular structures are small, tangled, and
-incredibly numerous. Therefore (<a href="#REF_B_24">Young (24)</a>), “ ... physiologists
-hope that after having learned a lot about nerve-impulses in the nerves they
-will be able to go on to study how these impulses interact when they
-reach the brain. [But], we must not assume that we shall understand
-the brain only in the terms we have learned to use for the nerves.
-The function of nerves is to carry impulses—like telegraph wires.
-The functions of brains is something else.” But, confronted with such
-awesome experimental difficulties, with no comprehensive mathematical
-theory in sight, we are largely limited otherwise to verbal discourses,
-rationales and theorizing, a hopelessly clumsy tool for the development
-of an adequate understanding of brain function. A little over ten years
-ago <a href="#REF_B_19">Sperry (19)</a> said, “Present day science is quite at
-a loss even to begin to describe the neural events involved in the simplest form
-of mental activity.” This situation has not changed much today. The
-development, study, and understanding of complex high-density cellular
-<span class="pagenum" id="Page_21">[Pg 21]</span>
-structures which incorporate characteristics of both the Lillie and
-Pask models may, it is hoped, alleviate this situation. There would
-also be fairly obvious technological applications for such techniques
-if highly developed and which, more than any other consideration, has
-prompted support for this work.</p>
-
-<p>Experiments to date have been devised which demonstrate the following
-basic physical functional characteristics:</p>
-
-<div class="blockquot">
-<p class="neg-indent">(1) Control of bulk resistivity of electrolytes
-containing closely-packed, poorly-conducting pellets</p>
-
-<p class="neg-indent">(2) Circulation of regenerative waves on closed
-loops</p>
-
-<p class="neg-indent">(3) Strong coupling between isolated excitable
-sites</p>
-
-<p class="neg-indent">(4) Logically-complete wave interactions,
-including facilitation and annihilation</p>
-
-<p class="neg-indent">(5) Dendrite growth by electrodeposition in
-“closed” excitable systems</p>
-
-<p class="neg-indent">(6) Subthreshold distributed field effects,
-especially in locally-refractory regions.</p>
-</div>
-
-<p>In addition, our attention has necessarily been directed to various
-problems of general experimental technique and choice of materials,
-especially as related to stability, fast recovery and long life.
-However, in order to understand the possible significance of, and
-motivation for such experiments, some related modern concepts of
-neurophysiology, histology and psychology will be reviewed very
-briefly. These concepts are, respectively:</p>
-
-<ul class="index">
-<li class="isub1">(1) Cellular structure in the central nervous system</li>
-<li class="isub1">(2) Short-term or “ephemeral” memory</li>
-<li class="isub1">(3) The synapse</li>
-<li class="isub1">(4) Inhibition</li>
-<li class="isub1">(5) Long-term memory traces or engram</li>
-<li class="isub1">(6) Spatially-diffuse temporal association and learning.</li>
-</ul>
-
-<h3>SOME CONTEMPORARY CONCEPTS</h3>
-
-<p>Since we are attempting to duplicate processes other than chemical,
-per se, we will forego any reference to the extensive literature of
-neurochemistry. It should not be surprising though if, at the neglect
-of the fundamental biological processes of growth, reproduction and
-metabolism, it proves possible to imitate some learning mechanisms with
-<span class="pagenum" id="Page_22">[Pg 22]</span>
-grossly less complex molecular structures. There is also much talk of
-chemical versus electrical theories and mechanisms in neurophysiology.
-The distinction, when it can be made, seems to hinge on the question
-of the scale of size of significant interactions. Thus, “chemical”
-interactions presumably take place at molecular distances, possibly as
-a result of or subsequent to a certain amount of thermal diffusion.
-“Electrical” interactions, on the other hand, are generally understood
-to imply longer range or larger scale macroscopic fields.</p>
-
-<h4>1. Cellular Structure</h4>
-
-<p>The human brain contains approximately 10¹⁰ neurons to which the
-neuron theory assigns the primary role in central nervous activity.
-These cells occupy, however, a relatively small fraction of the total
-volume. There are, for example, approximately 10 times that number of
-neuroglia, cells of relatively indeterminate function. Each neuron
-(consisting of cell body, dendrites and, sometimes, an axon) comes into
-close contact with the dendrites of other neurones at some thousands
-of places, these synapses and “ephapses” being spaced approximately 5μ
-apart <a href="#REF_B_1">(1)</a>. The total number of such apparent junctions is
-therefore of the order of 10¹³. In spite of infinite fine-structure variations
-when viewed with slightly blurred vision, the cellular structure of
-the brain is remarkably homogeneous. In the cortex, at least, the
-extensions of most cells are relatively short, and when the cortex is
-at rest, it appears from the large EEG alpha-rhythms that large numbers
-of cells beat together in unison. Quoting again from Sperry, “In short,
-current brain theory encourages us to try to correlate our subjective
-psychic experience with the activity of relatively homogeneous nerve
-cell units conducting essentially homogeneous impulses, through roughly
-homogeneous cerebral tissue.”</p>
-
-<h4>2. Short-Term Memory</h4>
-
-<p>A train of impulses simply travelling on a long fiber may, for
-example, be regarded as a short-term memory much in the same way as
-a delay line acts as a transient memory in a computer. A similar but
-slightly longer term memory may also be thought of to exist in the
-form of waves circulating in closed loops <a href="#REF_B_23">(23)</a>. In fact,
-it is almost universally held today that most significant memory occurs in two
-basic interrelated ways. First of all, such a short-term circulating,
-reverberatory or regenerative memory which, however, could not
-<span class="pagenum" id="Page_23">[Pg 23]</span>
-conceivably persist through such things as coma, anesthesia,
-concussion, extreme cold, deep sleep and convulsive seizures and
-thus, secondly, a long-term memory trace which must somehow reside
-in a semipermanent fine-structural change. As <a href="#REF_B_9">Hebb (9)</a> stated,
-“A reverbratory trace might cooperate with a structural change and carry
-the memory until the growth change is made.”</p>
-
-<h4>3. The Synapse</h4>
-
-<p>The current most highly regarded specific conception of the synapse
-is largely due to and has been best described by <a href="#REF_B_5">Eccles (5)</a>: “ ...
-the synaptic connections between nerve cells are the only functional
-connections of any significance. These synapses are of two types,
-excitatory and inhibitory, the former type tending to make nerve cells
-discharge impulses, the other to suppress the discharge. There is now
-convincing evidence that in vertebrate synapses each type operates
-through specific chemical transmitter substances ...”. In response to
-a presentation by <a href="#REF_B_10">Hebb (10)</a>, Eccles was quoted as saying,
-“One final point, and that is if there is electrical interaction, and we have seen
-from Dr. Estable’s work the complexity of connections, and we now know
-from the electronmicroscopists that there is no free space, only 200
-Å clefts, everywhere in the central nervous system, then everything
-should be electrically interacted with everything else. I think this is
-only electrical background noise and, that when we lift with specific
-chemical connections above that noise we get a significant operational
-system. I would say that there is electrical interaction but it is just
-a noise, a nuisance.” Eccles’ conclusions are primarily based on data
-obtained in the peripheral nervous system and the spinal cord. But
-there is overwhelming reason to expect that cellular interactions in
-the brain are an entirely different affair. For example, “The highest
-centres in the octopus, as in vertebrates and arthropods, contain many
-small neurons. This finding is such a commonplace, that we have perhaps
-failed in the past to make the fullest inquiry into its implications.
-Many of these small cells possess numerous processes, but no axon. It
-is difficult to see, therefore, that their function can be conductive
-in the ordinary sense. Most of our ideas about nervous functioning are
-based on the assumption that each neuron acts essentially as a link in
-some chain of conduction, but there is really no warrant for this in
-the case of cells with many short branches. Until we know more of the
-relations of these processes to each other in the neuropile it would
-be unwise to say more. It is possible that the effective part of the
-<span class="pagenum" id="Page_24">[Pg 24]</span>
-discharge of such cells is not as it is in conduction in long pathways,
-the internal circuit that returns through the same fiber, but the
-external circuit that enters other processes, ...” <a href="#REF_B_3">(3)</a>.</p>
-
-<h4>4. Inhibition</h4>
-
-<p>The inhibitory chemical transmitter substance postulated by Eccles
-has never been detected in spite of numerous efforts to do so. The
-mechanism(s) of inhibition is perhaps the key to the question of
-cellular interaction and, in one form or another, must be accounted for
-in any adequate theory.</p>
-
-<p>Other rather specific forms of excitation and inhibition interaction
-have been proposed at one time or another. Perhaps the best example is
-the polar neuron of <a href="#REF_B_8">Gesell (8)</a> and, more
-recently, <a href="#REF_B_18">Retzlaff (18)</a>. In such a
-concept, excitatory and inhibitory couplings differ basically
-because of a macroscopic structural difference at the cellular level;
-that is, various arrangements or orientation of intimate cellular
-structures give rise to either excitation or inhibition.</p>
-
-<h4>5. Long-Term Memory</h4>
-
-<p>Most modern theories of semipermanent structural change (or
-<i>engrams</i>, as they are sometimes called) look either to the
-molecular level or to the cellular level. Various specific locales for
-the engram have been suggested, including <a href="#REF_B_1">(1)</a> modifications of RNA
-molecular structure, <a href="#REF_B_2">(2)</a> changes of cell size, synapse area or
-dendrite extensions, <a href="#REF_B_3">(3)</a> neuropile modification, and
-<a href="#REF_B_4">(4)</a> local changes in the cell membrane.
-There is, in fact, rather direct evidence of the growth
-of neurons or their dendrites with use and the diminution or atrophy
-of dendrites with disuse. The apical dendrite of pyramidal neurones
-becomes thicker and more twisted with continuing activity, nerve fibers
-swell when active, sprout additional branches (at least in the spinal
-cord) and presumably increase the size and number of their terminal
-knobs. As pointed out by <a href="#REF_B_11">Konorski (11)</a>, the morphological
-conception of plasticity according to which plastic changes would be related to
-the formation and multiplication of new synaptic junctions goes back at
-least as far as Ramon y Cajal in 1904. Whatever the substrate of the
-memory trace, it is, at least in adults, remarkably immune to extensive
-brain damage and as <a href="#REF_B_24">Young (24)</a> has said: “ ... this question
-of the nature of the memory trace is one of the most obscure and disputed in
-the whole of biology.”
-<span class="pagenum" id="Page_25">[Pg 25]</span></p>
-
-<h4>6. Field Effects and Learning</h4>
-
-<p>First, from <a href="#REF_B_3">Boycott and Young (3)</a>, “The current conception,
-on which most discussions of learning still concentrate, is that the nervous
-system consists essentially of an aggregate of chains of conductors,
-linked at key points by synapses. This reflex conception, springing
-probably from Cartesian theory and method, has no doubt proved of
-outstanding value in helping us to analyse the actions of the spinal
-cord, but it can be argued that it has actually obstructed the
-development of understanding of cerebral function.”</p>
-
-<p>Most observable evidence of learning and memory is extremely complex
-and its interpretation full of traps. Learning in its broadest sense
-might be detected as a semipermanent change of behavior pattern brought
-about as a result of experience. Within that kind of definition, we
-can surely identify several distinctly different types of learning,
-presumably with distinctly different kinds of mechanisms associated
-with each one. But, if we are to stick by our definition of a condition
-of semipermanent change of behavior as a criterion for learning, then
-we may also be misled into considering the development of a neurosis,
-for example, as learning, or even a deep coma as learning.</p>
-
-<p>When we come to consider field effects, current theories tend to get
-fairly obscure, but there seems to be an almost universal recognition
-of the fact that such fields are significant. For example, <a href="#REF_B_16">Morrell (16)</a>
-says in his review of electrophysiological contributions to the
-neural basis of learning, “A growing body of knowledge (see reviews
-by Purpura, Grundfest, and Bishop) suggests that the most significant
-integrative work of the central nervous system is carried on in
-graded response elements—elements in which the degree of reaction
-depends upon stimulus intensity and is not all-or-none, which have no
-refractory period and in which continuously varying potential changes
-of either sign occur and mix and algebraically sum.” <a href="#REF_B_7">Gerard (7)</a>
-also makes a number of general comments along these lines. “These attributes
-of a given cell are, in turn, normally controlled by impulses arising
-from other regions, by fields surrounding them—both electric and
-chemical—electric and chemical fields can strongly influence the
-interaction of neurones. This has been amply expounded in the case
-of the electric fields.”</p>
-
-<p>Learning situations involving “punishment” and “reward” or,
-subjectively, “pain” and “pleasure” may very likely be associated with
-transient but structurally widespread field effects. States of distress
-<span class="pagenum" id="Page_26">[Pg 26]</span>
-and of success seem to exert a lasting influence on behavior only in
-relation to <i>simultaneous</i> sensory events or, better yet, sensory
-events just immediately <i>preceding</i> in time. For example, the
-“anticipatory” nature of a conditioned reflex has been widely noted
-<a href="#REF_B_21">(21)</a>. From a structural point of view, it is as if recently
-active sites regardless of location or function were especially sensitive to
-extensive fields. There is a known inherent electrical property of both
-nerve membrane and passive iron surface that could hold the answer to
-this mechanism of spatially-diffuse temporal association; namely, the
-surface resistance drops to less than 1 per cent of its resting value
-during the refractory period which immediately follows activation.</p>
-
-<h3>EXPERIMENTAL TECHNIQUE</h3>
-
-<p>In almost all experiments, the basic signal-energy mechanism employed
-has been essentially that one studied most extensively by <a href="#REF_B_12">Lillie (12)</a>,
-<a href="#REF_B_2">Bonhoeffer (2)</a>, <a href="#REF_B_22">Yamagiwa (22)</a>,
-<a href="#REF_B_14">Matumoto and Goto (14)</a> and others,
-<i>i.e.</i>, activation, impulse propagation and recovery on the
-normally passive surface of a piece of iron immersed in nitric acid or
-of cobalt in chromic acid <a href="#REF_B_20">(20)</a>. The iron we have used most
-frequently is of about 99.99% purity, which gives performance more consistent
-than but similar to that obtained using cleaned “coat-hanger” wires.
-The acid used most frequently by us is about 53-55% aqueous solution
-by weight, substantially more dilute than that predominantly used by
-previous investigators. The most frequently reported concentration has
-been 68-70%, a solution which is quite stable and, hence, much easier
-to work with in open containers than the weaker solutions, results in
-very fast waves but gives, at room temperatures, a very long refractory
-period (typically, 15 minutes). A noble metal (such as silver, gold
-or platinum) placed in contact with the surface of the iron has a
-stabilizing effect <a href="#REF_B_14">(14)</a> presumably through the action of
-local currents and provides a simple and useful technique whereby, with dilution,
-both stability and fast recovery (1 second) can be achieved in simple
-demonstrations and experiments.</p>
-
-<p>Experiments involving the growth by electrodeposition and study of
-metallic dendrites are done with an eye toward electrical, physical
-and chemical compatibility with the energy-producing system outlined
-above. Best results to date (from the standpoints of stability,
-non-reactivity, and morphological similarity to neurological
-structures) have been obtained by dissolving various amounts of gold
-chloride salt in 53-55% HNO₃.
-<span class="pagenum" id="Page_27">[Pg 27]</span></p>
-
-<p>An apparatus has been devised and assembled for the purpose of
-containing and controlling our primary experiments. (<a href="#FIG_1B">See Figure 1</a>).
-Its two major components are a test chamber (on the left in <a href="#FIG_1B">Figure 1</a>)
-and a fluid exchanger (on the right). In normal operation the
-test chamber, which is very rigid and well sealed after placing the
-experimental assembly inside, is completely filled with electrolyte
-(or, initially, an inert fluid) to the exclusion of all air pockets
-and bubbles. Thus encapsulated, it is possible to perform experiments
-which would otherwise be impossible due to instability. The instability
-which plagues such experiments is manifested in copious generation
-of bubbles on and subsequent rapid disintegration of all “excitable”
-material (<i>i.e.</i>, iron). Preliminary experiments indicated that
-such “bubble instability” could be suppressed by constraining the
-volume available to expansion. In particular, response and recovery
-times can now be decreased substantially and work can proceed with
-complex systems of interest such as aggregates containing many small
-iron pellets.</p>
-
-<p>The test chamber is provided with a heater (and thermostatic control)
-which makes possible electrochemical impulse response and recovery
-times comparable to those of the nervous system (1 to 10 msec). The
-fluid-exchanger is so arranged that fluid in the test chamber can be
-arbitrarily changed or renewed by exchange within a rigid, sealed,
-completely liquid-filled (“isochoric”) loop. Thus, stability can
-be maintained for long periods of time and over a wide variety of
-investigative or operating conditions.</p>
-
-<p>Most of the parts of this apparatus are made of stainless steel and
-are sealed with polyethylene and teflon. There is a small quartz
-observation window on the test chamber, two small lighting ports, a
-pressure transducer, thermocouple, screw-and-piston pressure actuator
-and umbilical connector for experimental electrical inputs and outputs.</p>
-
-<h3>BASIC EXPERIMENTS</h3>
-
-<p>The basic types of experiments described in the following sections
-are numbered for comparison to correspond roughly to related
-neurophysiological concepts summarized in the previous section.</p>
-
-<h4>1. Cellular Structure</h4>
-
-<p>The primary object of our research is the control and determination of
-dynamic behavior in response to electrical stimulation in close-packed
-aggregates of small pellets submerged in electrolyte. Typically, the
-aggregate contains (among other things) iron and the electrolyte
-contains nitric acid, this combination making possible the propagation
-of electrochemical surface waves of excitation through the body of
-the aggregate similar to those of the Lillie iron-wire nerve model.
-The iron pellets are imbedded in and supported by a matrix of small
-dielectric (such as glass) pellets. Furthermore, with the addition
-of soluble salts of various noble metals to the electrolyte, long
-interstitial dendritic or fibrous structures of the second metal can
-be formed whose length and distribution change by electrodeposition in
-response to either internal or externally generated fields.
-<span class="pagenum" id="Page_28">[Pg 28]</span></p>
-
-<div class="figcenter">
- <img id="FIG_1B" src="images/i_035.jpg" alt="" width="300" height="714" />
- <p class="f120 space-below1">Figure 1—Test chamber and<br /> fluid exchanger</p>
-</div>
-
-<p><span class="pagenum" id="Page_29">[Pg 29]</span>
-Coupling between isolated excitable (iron) sites is greatly affected
-by the fine structure and effective bulk resistivity of the glass and
-fluid medium which supports and fills the space between such sites.
-In general (<a href="#SECT_3">see Section 3, following</a>) it is necessary, to
-promote strong coupling between small structures, to impede the “short-circuit”
-return flow of current from an active or excited surface, through
-the electrolyte and back through the dendritic structure attached to
-the same excitable site. This calls for control (increase) of the
-bulk resistivity, preferably by means specifically independent of
-electrolyte composition, which relates to and affects surface phenomena
-such as recovery (<i>i.e.</i>, the “refractory” period). <a href="#FIG_2B">Figure 2</a>
-illustrates the way in which this is being done, <i>i.e.</i>, by
-appropriate choice of particle size distributions. The case illustrated
-shows the approximate proper volume ratios for maximum resistivity in a
-two-size-phase random mixture of spheres.</p>
-
-<h4>2. Regenerative Loops</h4>
-
-<p><a href="#FIG_3B">Figure 3</a> shows an iron loop (about 2-inch diameter) wrapped
-with a silver wire helix which is quite stable in 53-55% acid and which
-will easily support a circulating pattern of three impulses. For
-demonstration, unilateral waves can be generated by first touching the
-iron with a piece of zinc (which produces two oppositely travelling
-waves) and then blocking one of them with a piece of platinum or a
-small platinum screen attached to the end of a stick or wand. Carbon
-blocks may also be used for this purpose.</p>
-
-<p>The smallest regenerative or reverberatory loop which we are at present
-able to devise is about 1 mm in diameter. Multiple waves, as expected,
-produce stable patterns in which all impulses are equally spaced. This
-phenomenon can be related to the slightly slower speed characteristic
-of the relative refractory period as compared with a more fully
-recovered zone.
-<span class="pagenum" id="Page_30">[Pg 30]</span></p>
-
-<div class="figcenter">
- <img id="FIG_2B" src="images/i_037a.jpg" alt="" width="600" height="247" />
- <img src="images/i_037b.jpg" alt="" width="600" height="275" />
- <p class="f120 space-below2">Figure 2—Conductivity control—mixed pellet-size aggregates</p>
-</div>
-<p><span class="pagenum" id="Page_31">[Pg 31]</span></p>
-<div class="figcenter">
- <img id="FIG_3B" src="images/i_038.jpg" alt="" width="500" height="510" />
- <p class="f120 space-below2">Figure 3—Regenerative or reverberatory loop</p>
-</div>
-
-<h4 id="SECT_3">3. Strong Coupling</h4>
-
-<p>If two touching pieces of iron are placed in a bath of nitric acid, a
-wave generated on one will ordinarily spread to the other. As is to be
-expected, a similar result is obtained if the two pieces are connected
-through an external conducting wire. However, if they are isolated,
-strong coupling does not ordinarily occur, especially if the elements
-are small in comparison with a “critical size,” σ/ρ where σ is the
-surface resistivity of passive iron surface (in Ω-cm²) and ρ is the
-volume resistivity of the acid (in Ω-cm). A simple and informative
-structure which demonstrates the essential conditions for strong
-electrical coupling between isolated elements of very small size may
-be constructed as shown in <a href="#FIG_4B">Figure 4</a>. The dielectric barrier
-insures that charge transfer through one dipole must be accompanied by an
-equal and opposite transfer through the surfaces of the other dipole.
-If the “inexcitable” silver tails have sufficiently high conductance
-(<i>i.e.</i>, sufficiently large surface area, hence preferably,
-dendrites), strong coupling will occur, just as though the cores of the
-two pieces of iron were connected with a solid conducting wire.
-<span class="pagenum" id="Page_32">[Pg 32]</span></p>
-
-<div class="figcenter">
- <img id="FIG_4B" src="images/i_039a.jpg" alt="" width="600" height="348" />
- <p class="f120 space-below2">Figure 4</p>
-
- <img id="FIG_5B" src="images/i_039b.jpg" alt="" width="400" height="521" />
- <p class="f120 space-below2">Figure 5—Electrochemical excitatory-inhibitory<br />
- interaction cell</p>
-</div>
-<p><span class="pagenum" id="Page_33">[Pg 33]</span></p>
-
-<h4>4. Inhibitory Coupling</h4>
-
-<p>If a third “dipole” is inserted through the dielectric membrane in
-the opposite direction, then excitation of this isolated element
-tends to inhibit the response which would otherwise be elicited by
-excitation of one of the parallel dipoles. <a href="#FIG_5B">Figure 5</a> shows the
-first such “logically-complete” interaction cell successfully constructed and
-demonstrated. It may be said to behave as an elementary McCulloch-Pitts
-neuron <a href="#REF_B_15">(15)</a>. Further analysis shows that similar structures
-incorporating many dipoles (both excitatory and inhibitory) can be made
-to behave as general “linear decision functions” in which all input
-weights are approximately proportional to the total size or length of
-their corresponding attached dendritic structures.</p>
-
-<h4>5. Dendrite Growth</h4>
-
-<p><a href="#FIG_6B">Figure 6</a> shows a sample gold dendrite grown by electrodeposition
-(actual size, about 1 mm) from a 54% nitric acid solution to which gold
-chloride was added. When such a dendrite is attached to a piece of
-iron (both submerged), activation of the excitable element produces a
-field in such a direction as to promote further growth of the dendritic
-structure. Thus, if gold chloride is added to the solution used in
-the elementary interaction cells described above, all input influence
-“weights” tend to increase with use and, hence, produce a plasticity of
-function.</p>
-
-<h4>6. Field Effects in Locally-Refractory Regions</h4>
-
-<p>Our measurements indicate that, during the refractory period following
-excitation, the surface resistance of iron in nitric acid drops to
-substantially less than 1% of its resting value in a manner reminiscent
-of nerve membranes <a href="#REF_B_4">(4)</a>. Thus, if a distributed or gross
-field exists at any time throughout a complex cellular aggregate, concomitant
-current densities in locally-refractive regions will be substantially
-higher than elsewhere and, if conditions appropriate to dendrite
-growth exist (as described above) growth rates in such regions will
-also be substantially higher than elsewhere. It would appear that, as
-a result, recently active functional couplings (in contrast to those
-not associated with recent neural activity) should be significantly
-altered by widely distributed fields or massive peripheral shocks. This
-mechanism might thus explain the apparent ability of the brain to form
-specific temporal associations in response to spatially-diffuse effects
-such as are generated, for example, by the pain receptors.
-<span class="pagenum" id="Page_34">[Pg 34]</span></p>
-
-<div class="figcontainer">
- <div class="figsub">
- <img id="FIG_6B" src="images/i_041a.jpg" alt="" width="300" height="317" />
- </div>
- <div class="figsub">
- <img src="images/i_041b.jpg" alt="" width="250" height="328" />
- </div>
- <p class="f120 space-below2"><b>(a)</b></p>
- <img src="images/i_041c.jpg" alt="" width="400" height="378" />
- <p class="f120 space-below2"><b>(b)</b></p>
- <div class="blockquot">
- <p class="f120">Figure 6—Dendritic structures, living and non-living. <b>(a)</b> Cat dendrite
- trees (from Bok, “Histonomy of the Cerebral Cortex,” Elsevier, 1959);
- <b>(b)</b> Electrodeposited gold dendrite tree.</p>
- </div>
-</div>
-<p class="space-below2"><span class="pagenum" id="Page_35">[Pg 35]</span></p>
-
-<h3>SUMMARY</h3>
-
-<p>An attempt is being made to develop meaningful electrochemical
-model techniques which may contribute toward a clearer understanding
-of cortical function. Two basic phenomena are simultaneously employed
-which are variants of (1) the Lillie iron-wire nerve model, and (2)
-growth of metallic dendrites by electrodeposition. These phenomena
-are being induced particularly within dense cellular aggregates of
-various materials whose interstitial spaces are flooded with liquid
-electrolyte.</p>
-
-<p class="f120 space-above1"><b>REFERENCES</b></p>
-<table border="0" cellspacing="0" summary="REFERENCES" cellpadding="2" >
- <tbody><tr>
- <td id="REF_B_1" class="tdr">1.</td>
- <td class="tdl_ws1">Bok, S. T.,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">“Histonomy of the Cerebral Cortex,”</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws2">Amsterdam, London:Elsevier Publishing Co., New York:Princeton, 1959</td>
- </tr><tr>
- <td id="REF_B_2" class="tdr">2.</td>
- <td class="tdl_ws1">Bonhoeffer, K. F.,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">“Activation of Passive Iron as a Model for the Excitation of Nerve,”</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws2"><i>J. Gen. Physiol.</i> <b>32</b>:69-91 (1948).</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1"><p class="no-indent">This paper summarizes work carried out during 1941-1946
- at the University of Leipzig, and published during the war years in German periodicals.</p></td>
- </tr><tr>
- <td id="REF_B_3" class="tdr">3.</td>
- <td class="tdl_ws1">Boycott, B. B., and Young, J. Z.,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">“The Comparative Study of Learning,”</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">S. E. B. Symposia, No. IV</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">“Physiological Mechanisms in Animal Behavior,”</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws2">Cambridge: University Press, USA:Academic Press, Inc., 1950</td>
- </tr><tr>
- <td id="REF_B_4" class="tdr">4.</td>
- <td class="tdl_ws1">Cole, K. S., and Curtis, H. J.,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">“Electric Impedance of the Squid Giant Axon During Activity,”</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws2"><i>J. Gen. Physiol.</i> <b>22</b>:649-670 (1939)</td>
- </tr><tr>
- <td id="REF_B_5" class="tdr">5.</td>
- <td class="tdl_ws1">Eccles, J. C.,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">“The Effects of Use and Disuse of Synaptic Function,”</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">“Brain Mechanisms and Learning—A Symposium,”</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1"><p class="no-indent">organized by the Council for International Organizations of
- Medical Science, Oxford:Blackwell Scientific Publications, 1961</p></td>
- </tr><tr>
- <td id="REF_B_6" class="tdr">6.</td>
- <td class="tdl_ws1">Franck, U. F.,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">“Models for Biological Excitation Processes,”</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">“Progress in Biophysics and Biophysical Chemistry,”</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">J. A. V. Butler, ed., London and New York:Pergamon Press,
- pp. 171-206, 1956</td>
- </tr><tr>
- <td id="REF_B_7" class="tdr">7.</td>
- <td class="tdl_ws1">Gerard, R. W.,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">“Biological Roots of Psychiatry,”</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws2"><i>Science</i> <b>122 (No. 3162)</b>:225-230 (1955)</td>
- </tr><tr>
- <td id="REF_B_8" class="tdr">8.</td>
- <td class="tdl_ws1">Gesell, R.,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">“A Neurophysiological Interpretation of the Respiratory Act,”</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws2"><i>Ergedn. Physiol.</i> <b>43:</b>477-639 (1940)</td>
- </tr><tr>
- <td id="REF_B_9" class="tdr">9.</td>
- <td class="tdl_ws1">Hebb, D. O.,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">“The Organization of Behavior, A Neuropsychological Theory,”</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws2">New York:John Wiley and Sons, 1949</td>
- </tr><tr>
- <td id="REF_B_10" class="tdr">10.</td>
- <td class="tdl_ws1">Hebb, D. O.,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">“Distinctive Features of Learning in the Higher Animal,”</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">“Brain Mechanisms and Learning—A Symposium,”</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1"><p class="no-indent">organized by the Council for International Organizations of
- Medical Science, Oxford:Blackwell Scientific Publications, 1961</p></td>
- </tr><tr>
- <td id="REF_B_11" class="tdr">11.</td>
- <td class="tdl_ws1">Konorski, J.,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">“Conditioned Reflexes and Neuron Organization,”</td>
- </tr><tr>
- <td class="tdr"><span class="pagenum" id="Page_36">[Pg 36]</span></td>
- <td class="tdl_ws2">Cambridge:Cambridge University Press, 1948</td>
- </tr><tr>
- <td id="REF_B_12" class="tdr">12.</td>
- <td class="tdl_ws1">Lillie, R. S.,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">“Factors Affecting the Transmission and Recovery
- in the Passive Iron Nerve Model,”</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws2"><i>J. Gen. Physiol.</i> <b>4</b>:473 (1925)</td>
- </tr><tr>
- <td id="REF_B_13" class="tdr">13.</td>
- <td class="tdl_ws1">Lillie, R. S.,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws2"><i>Biol. Rev.</i> <b>16</b>:216 (1936)</td>
- </tr><tr>
- <td id="REF_B_14" class="tdr">14.</td>
- <td class="tdl_ws1">Matumoto, M., and Goto, K.,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">“A New Type of Nerve Conduction Model,”</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws2"><i>The Gurma Journal of Medical Sciences</i> <b>4(No. 1)</b> (1955)</td>
- </tr><tr>
- <td id="REF_B_15" class="tdr">15.</td>
- <td class="tdl_ws1">McCulloch, W. S., and Pitts, W.,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">“A Logical Calculus of the Ideas Immanent in Nervous Activity,”</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws2"><i>Bulletin of Mathematical Biophysics</i> <b>5</b>:115-133 (1943)</td>
- </tr><tr>
- <td id="REF_B_16" class="tdr">16.</td>
- <td class="tdl_ws1">Morrell, F.,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">“Electrophysiological Contributions to the
- Neural Basis of Learning,”</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws2"><i>Physiological Reviews</i> <b>41(No. 3)</b> (1961)</td>
- </tr><tr>
- <td id="REF_B_17" class="tdr">17.</td>
- <td class="tdl_ws1">Pask, G.,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">“The Growth Process Inside the Cybernetic Machine,”</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws2"><i>Proc. 2nd Congress International Association Cybernetics</i>,
- Gauthier-Villars, Paris:Namur, 1958</td>
- </tr><tr>
- <td id="REF_B_18" class="tdr">18.</td>
- <td class="tdl_ws1">Retzlaff, E.,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">“Neurohistological Basis for the Functioning
- of Paired Half-Centers,”</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws2"><i>J. Comp. Neurology</i> <b>101</b>:407-443 (1954)</td>
- </tr><tr>
- <td id="REF_B_19" class="tdr">19.</td>
- <td class="tdl_ws1">Sperry, R. W.,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">“Neurology and the Mind-Brain Problem,”</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws2"><i>Amer. Scientist</i> <b>40(No. 2)</b>: 291-312 (1952)</td>
- </tr><tr>
- <td id="REF_B_20" class="tdr">20.</td>
- <td class="tdl_ws1">Tasaki, I., and Bak, A. F.,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws2"><i>J. Gen. Physiol.</i> <b>42</b>:899 (1959)</td>
- </tr><tr>
- <td id="REF_B_21" class="tdr">21.</td>
- <td class="tdl_ws1">Thorpe, W. H.,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">“The Concepts of Learning and Their Relation
- to Those of Instinct,”</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws2">S. E. B. Symposia, No. IV,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">“Physiological Mechanisms in Animal Behavior,”</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws2">Cambridge:University Press, USA:Academic Press, Inc., 1950</td>
- </tr><tr>
- <td id="REF_B_22" class="tdr">22.</td>
- <td class="tdl_ws1">Yamagiwa, K.,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">“The Interaction in Various Manifestations
- (Observations on Lillie’s Nerve Model),”</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws2"><i>Jap. J. Physiol.</i> <b>1</b>:40-54 (1950)</td>
- </tr><tr>
- <td id="REF_B_23" class="tdr">23.</td>
- <td class="tdl_ws1">Young, J. Z.,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">“The Evolution of the Nervous System and of the
- Relationship of Organism and Environment,”</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">G. R. de Beer, ed.,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">“Evolution,”</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws2">Oxford:Clarendon Press, pp. 179-204, 1938</td>
- </tr><tr>
- <td id="REF_B_24" class="tdr">24.</td>
- <td class="tdl_ws1">Young, J. Z.,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">“Doubt and Certainty in Science, A Biologist’s
- Reflections on the Brain,”</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws2">New York:Oxford Press, 1951</td>
- </tr>
- </tbody>
-</table>
-<hr class="chap x-ebookmaker-drop" />
-
-<div class="chapter">
-<p><span class="pagenum" id="Page_37">[Pg 37]</span></p>
-<h2 class="nobreak">Multi-Layer Learning Networks</h2>
-</div>
-
-<p class="f120"><b><span class="smcap">R. A. STAFFORD</span></b></p>
-
-<p class="center space-below1"><i>Philco Corp., Aeronutronic Division<br />
- Newport Beach, California</i></p>
-
-<h3>INTRODUCTION</h3>
-
-<p>This paper is concerned with the problem of designing a network of
-linear threshold elements capable of efficiently adapting its various
-sets of weights so as to produce a prescribed input-output relation.
-It is to accomplish this adaptation by being repetitively presented
-with the various inputs along with the corresponding desired outputs.
-We will not be concerned here with the further requirement of various
-kinds of ability to “generalize”—<i>i.e.</i>, to tend to give correct
-outputs for inputs that have not previously occurred when they are
-similar in some transformed sense to other inputs that have occurred.</p>
-
-<p>In putting forth a model for such an adapting or “learning” network, a
-requirement is laid down that the complexity of the adaption process
-in terms of interconnections among elements needed for producing
-appropriate weight changes, should not greatly exceed that already
-required to produce outputs from inputs with a static set of weights.
-In fact, it has been found possible to use the output-from-input
-computing capacity of the network to help choose proper weight changes
-by observing the effect on the output of a variety of possible weight
-changes.</p>
-
-<p>No attempt is made here to defend the proposed network model on
-theoretical grounds since no effective theory is known at present.
-Instead, the plausibility of the various aspects of the network model,
-combined with empirical results must suffice.</p>
-
-<h3>SINGLE ELEMENTS</h3>
-
-<p>To simplify the problem it is assumed that the network receives a set
-of two-valued inputs, x₁, x₂, ..., xₙ, and is required to produce only
-a single two-valued output, y. It is convenient to assign the numerical
-quantities +1 and -1 to the two values of each variable.</p>
-
-<p>The simplest network would consist of a single linear threshold element
-with a set of weights, c₀, c₁, c₂, ..., cₙ. These determine the
-<span class="pagenum" id="Page_38">[Pg 38]</span>
-output-input relation or function so that y is +1 or -1 according as
-the quantity, c₀ + c₁x₁ + c₂x₂ + ... + cₙxₙ, is positive or not,
-respectively. It is possible for such a single element to exhibit an
-adaptive behavior as follows. If, for a given set, x₁, x₂, ..., xₙ, the
-output, y, is correct, then make no changes to the weights. Otherwise
-change the weights according to the equations</p>
-
-<ul class="index fontsize_130">
-<li class="isub4">Δc₀ = y*</li>
-<li class="isub4">Δcᵢ = y*xᵢ,&emsp;i = 1,2, ...,n</li>
-
-<li class="isub4 space-above1"><small>where y* is the desired output.</small></li>
-</ul>
-
-<p>It has been shown by a number of people that the weights of such an
-element are assured of arriving at a set of values which produce the
-correct output-input relation after a sufficient number of errors,
-provided that such a set exists. An upper bound on the number of
-possible errors can be given which depends only on the initial weight
-values and the logical function to be learned. This does not, however,
-solve our network problem for two reasons.</p>
-
-<p>First, as the number, n, of inputs gets large, the number of errors
-to be expected for most functions which can be learned increases to
-unreasonable values. For example, for n = 6, most such functions
-result in 500 to 1000 errors compared to an average of 32 errors to be
-expected in a perfect learning device.</p>
-
-<p>Second, and more important, the fraction of those logical functions
-which can be generated in a single element becomes vanishingly small as
-n increases. For example, at n = 6 less than one in each three trillion
-logical functions is so obtainable.</p>
-
-<h3>NETWORKS OF ELEMENTS</h3>
-
-<p>It can be demonstrated that if a sufficiently large number of linear
-threshold elements is used, with the outputs of some being the inputs
-of others, then a final output can be produced which is any desired
-logical function of the inputs. The difficulty in such a network lies
-in the fact that we are no longer provided with a knowledge of the
-correct output for each element, but only for the final output. If the
-final output is incorrect there is no obvious way to determine which
-sets of weights should be altered.</p>
-
-<p>As a result of considerable study and experimentation at Aeronutronic,
-a network model has been evolved which, it is felt, will get around
-these difficulties. It consists of four basic features which will now
-be described.
-<span class="pagenum" id="Page_39">[Pg 39]</span></p>
-
-<h4>Positive Interconnecting Weights</h4>
-
-<p>It is proposed that all weights in elements attached to inputs which
-come from other elements in the network be restricted to positive
-values. (Weights attached to the original inputs to the network, of
-course, must be allowed to be of either sign.) The reason for such a
-restriction is this. If element 1 is an input to element 2 with weight
-c₁₂, element 2 to element 3 with weight c₂₃, <i>etc.</i>, then the sign
-of the product, c₁₂c₂₃ ..., gives the sense of the effect of a change
-in the output of element 1 on the final element in the chain (assuming
-this is the only such chain between the two elements). If these various
-weights were of either possible sign, then a decision as to whether or
-not to change the output in element 1 to help correct an error in the
-final element would involve all weights in the chain. Moreover, since
-there would in general be a multiplicity of such chains, the decision
-is rendered impossibly difficult.</p>
-
-<p>The above restriction removes this difficulty. If the output of any
-element in the network is changed, say, from -1 to +1, the effect on
-the final element, if it is affected at all, is in the same direction.</p>
-
-<p>It should be noted that this restriction does not seriously affect
-the logical capabilities of a network. In fact, if a certain logical
-function can be achieved in a network with the use of weights of
-unrestricted sign, then the same function can be generated in another
-network with only positive interconnecting weights and, at worst, twice
-the number of elements. In the worst case this is done by generating
-in the restricted network both the output and its complement for each
-element of the unrestricted network. (It is assumed that there are no
-loops in the network.)</p>
-
-<h4>A Variable Bias</h4>
-
-<p>The central problem in network learning is that of determining, for
-a given input, the set of elements whose outputs can be altered so
-as to correct the final element, and which will do the least amount
-of damage to previous adaptations to other inputs. Once this set has
-been determined, the incrementing rule given for a single element will
-apply in this case as well (subject to the restriction of leaving
-interconnecting weights positive), since the desired final output
-coincides with that desired for each of the elements to be changed
-(because of positive interconnecting weights).</p>
-
-<p>In the process of arriving at such a decision three factors need to be
-considered. Elements selected for change should tend to be those whose
-<span class="pagenum" id="Page_40">[Pg 40]</span>
-output would thereby be affected for a minimum number of other possible
-inputs. At the same time it should be ascertained that a change in
-each of the elements in question does indeed contribute significantly
-towards correcting the final output. Finally, a minimum number of such
-elements should be used.</p>
-
-<p>It would appear at first that this kind of decision is impossible to
-achieve if the complexity of the decision apparatus is kept comparable
-to that of the basic input-output network as mentioned earlier.
-However, in the method to be described it is felt that a reasonable
-approximation to these requirements will be achieved without an undue
-increase in complexity.</p>
-
-<p>It is assumed that in addition to its normal inputs, each element
-receives a variable input bias which we can call b. The output of every
-element should then be determined by the sign of the usual weighted
-sum of its inputs plus this bias quantity. This bias is to be the same
-for each element of the network. If b = 0 the network will behave
-as before. However, if b is increased gradually, various elements
-throughout the network will commence changing from -1 to +1, with one
-or a few changing at any one time as a rule. If b is decreased, the
-opposite will occur.</p>
-
-<p>Now suppose that for a given input the final output ought to be +1 but
-actually is -1. Assume that b is then raised so high that this final
-output is corrected. Then commence a gradual decline in b. Various
-elements may revert to -1, but until the final output does, no weights
-are changed. When the final output does revert to -1, it is due to an
-element’s having a sum (weighted sum plus bias) which just passed down
-through zero. This then caused a chain effect of changing elements
-up to the final element, but presumably this element is the only one
-possessing a zero sum. This can then be the signal for the weights on
-an element to change—a change of final output from right to wrong
-accompanied simultaneously by a zero sum in the element itself.</p>
-
-<p>After such a weight change, the final output will be correct once more
-and the bias can again proceed to fall. Before it reaches zero, this
-process may occur a number of times throughout the network. When the
-bias finally stands at zero with the final output correct, the network
-is ready for the next input. Of course if -1 is desired, the bias will
-change in the opposite direction.</p>
-
-<p>It is possible that extending the weight change process a little past
-the zero bias level may have beneficial results. This might increase
-the life expectancy of each learned input-output combination and
-thereby reduce the total number of errors. This is because the method
-<span class="pagenum" id="Page_41">[Pg 41]</span>
-used above can stop the weight correction process so that even though
-the final output is correct, some elements whose output are essential
-to the final output have sums close to zero, which are easily changed
-by subsequent weight changes.</p>
-
-<p>It will be noted that this method conforms to all three considerations
-mentioned previously. First, by furnishing each element the same bias,
-and by not changing weights until the final output becomes incorrect
-with dropping bias, there is a strong tendency to select elements
-which, with b = 0, would have sums close to zero. But the size of the
-sum in an element is a good measure of the amount of damage done to
-an element for other inputs if its current output is to be changed.
-Second, it is obvious that each element changed has had a demonstrable
-effect on the final output. Finally, there will be a clear tendency to
-change only a minimum of elements because changes never occur until the
-output clearly requires a change.</p>
-
-<p>On the other hand this method requires little more added complexity to
-the network than it already has. Each element requires a bias, an error
-signal, and the desired final output, these things being uniform for
-all elements in a network. Some external device must manipulate the
-bias properly, but this is a simple behavior depending only on an error
-signal and the desired final output—not on the state of individual
-elements in the network. What one has, then, is a network consisting
-of elements which are nearly autonomous as regards their decisions
-to change weights. Such a scheme appears to be the only way to avoid
-constructing a central weight-change decision apparatus of great
-complexity. This rather sophisticated decision is made possible by
-utilizing the computational capabilities the network already possesses
-in producing outputs from inputs.</p>
-
-<p>It should be noted here that this varying bias method requires that
-the variable bias be furnished to just those elements which have
-variable weights and to no others. Any fixed portion of the network,
-such as preliminary layers or final majority function for example,
-must operate independently of the variable bias. Otherwise, the final
-output may go from right to wrong as the bias moves towards zero and no
-variable-weight element be to blame. In such a case the network would
-be hung up.</p>
-
-<h4>Logical Redundancy in the Network</h4>
-
-<p>A third aspect of the network model is that for all the care taken in
-the previous steps, they will not suffice in settling quickly to a set
-<span class="pagenum" id="Page_42">[Pg 42]</span>
-of weights that will generate the required logical function unless
-there is a great multiplicity of ways in which this can be done. This
-is to say that a learning network needs to have an excess margin of
-weights and elements beyond the minimum required to generate the
-functions which are to be learned.</p>
-
-<p>This is analogous to the situation that prevails for a single element
-as regards the allowed range of values on its weights. It can be shown
-for example, that any function for n=6 that can be generated by a
-single element can be obtained with each weight restricted to the range
-of integer values -9,-8, ..., +9. Yet no modification of the stated
-weight change rule is known which restricts weight values to these and
-yet has any chance of ever being learned for most functions.</p>
-
-<h4>Fatigued Elements</h4>
-
-<p>It would appear from some of the preliminary results of network
-simulations that it may be useful to have elements become “fatigued”
-after undergoing an excessive number of weight changes. Experiments
-have been performed on simplifications of the model described so far
-which had the occasional result that a small number of elements came
-to a state where they received most of the weight increments, much
-to the detriment of the learning process. In such cases the network
-behaves as if it were composed of many fewer adjustable elements. In a
-sense this is asking each element to maintain a record of the data it
-is being asked to store so that it does not attempt to exceed its own
-information capacity.</p>
-
-<p>It is not certain just how this fatigue factor should enter in the
-element’s actions, but if it is to be compatible with the variable bias
-method, this fatigue factor must enter into the element’s response to
-a changing bias. Once an element changes state with zero sum at the
-same time that the final output becomes wrong, incrementing must occur
-if the method is to work. Hence a “fatigued” element must respond less
-energetically to a change of bias, perhaps with a kind of variable
-factor to be multiplied by the bias term.</p>
-
-<h3>NETWORK STRUCTURE</h3>
-
-<p>It is felt that the problem of selecting the structure of
-interconnections for a network is intimately connected to the
-previously mentioned problem of generalization. Presumably a given
-type of generalization can be obtained by providing appropriate fixed
-<span class="pagenum" id="Page_43">[Pg 43]</span>
-portions of the network and an appropriate interconnection structure
-for the variable portion. However, for very large networks, it is
-undoubtedly necessary to restrict the complexity so that it can be
-specified by relatively simple rules. Since very little is known about
-this quite important problem, no further discussion will be attempted here.</p>
-
-<h3>COMPUTER SIMULATION RESULTS</h3>
-
-<p>A computer simulation of some of the network features previously
-described has been made on an IBM 7090. Networks with an excess of
-elements and with only positive interconnecting weights were used.
-However, in place of the variable bias method, a simple choice of the
-element of sum closest to, and on the wrong side of, zero was made
-without regard to the effectiveness of the element in correcting the
-final output. No fatigue factors were used.</p>
-
-<p>The results of these simulations are very encouraging, but at the same
-time indicate the need for the more sophisticated methods. No attempt
-will be made here to describe the results completely.</p>
-
-<p>In one series of learning experiments, a 22-element network was used
-which had three layers, 10 elements on the first, 11 on the second, and
-1 on the third. The single element on the third was the final output,
-and was a fixed majority function of the 11 elements in the second
-layer. These in turn each received inputs from each of the 10 on the
-first layer and from each of the 6 basic inputs. The 10 on the first
-layer each received only the 6 basic inputs. A set of four logical
-functions, A, B, C, and D, was used. Function A was actually a linear
-threshold function which could be generated by the weights 8, 7, 6, 5,
-4, 3, 2, functions B and C were chosen by randomly filling in a truth
-table, while D was the parity function.</p>
-
-<table id="TABLE-1" border="0" cellspacing="0" summary="Table I" cellpadding="2" >
- <caption><big><b>TABLE I</b></big></caption>
- <thead><tr>
- <th class="tdc bb2" colspan="8">&nbsp;</th>
- </tr><tr>
- <th class="tdc br" colspan="2"><big>A</big></th>
- <th class="tdc br" colspan="2"><big>B</big></th>
- <th class="tdc br" colspan="2"><big>C</big></th>
- <th class="tdc" colspan="2"><big>D</big></th>
- </tr><tr>
- <th class="tdc bb">&emsp;r&emsp;</th> <th class="tdc bb br">e&nbsp;</th>
- <th class="tdc bb">&emsp;r&emsp;</th> <th class="tdc bb br">e</th>
- <th class="tdc bb">&emsp;r&emsp;</th> <th class="tdc bb br">e</th>
- <th class="tdc bb">&emsp;r&emsp;</th> <th class="tdc bb">e</th>
- </tr>
- </thead>
- <tbody><tr>
- <td class="tdc">5</td> <td class="tdc br">54&nbsp;</td>
- <td class="tdc">8</td> <td class="tdc br">100&#8199;</td>
- <td class="tdc">11</td> <td class="tdc br">101&#8199;</td>
- <td class="tdc">4</td> <td class="tdc">52&nbsp;</td>
- </tr><tr>
- <td class="tdc">4</td> <td class="tdc br">37&nbsp;</td>
- <td class="tdc">9</td> <td class="tdc br">85&nbsp;</td>
- <td class="tdc">4</td> <td class="tdc br">60&nbsp;</td>
- <td class="tdc">5</td> <td class="tdc">62&nbsp;</td>
- </tr><tr>
- <td class="tdc">4</td> <td class="tdc br">44&nbsp;</td>
- <td class="tdc">6</td> <td class="tdc br">72&nbsp;</td>
- <td class="tdc">9</td> <td class="tdc br">85&nbsp;</td>
- <td class="tdc">6</td> <td class="tdc">56&nbsp;</td>
- </tr><tr>
- <td class="tdc bt2" colspan="8">&nbsp;</td>
- </tr>
- </tbody>
-</table>
-
-<p><span class="pagenum" id="Page_44">[Pg 44]</span>
-Table I gives the results of one series of runs with these functions
-and this network, starting with various random initial weights. The
-quantity, r, is the number of complete passes through the 64-entry
-truth table before the function was completely learned, while e is
-the total number of errors made. In evaluating the results it should
-be noted that an ideal learning device would make an average of 32
-errors altogether on each run. The totals recorded in these runs are
-agreeably close to this ideal. As expected, the linear threshold
-function is the easiest to learn, but it is surprising that the
-parity function was substantially easier than the two randomly chosen
-functions. <a href="#TABLE-2">Table II</a> gives a chastening result of the same
-experiment with all interconnecting weights removed except that the final element
-is a fixed majority function of the other 21 elements. Thus there was
-adaptation on one layer only. As can be seen <a href="#TABLE-1">Table I</a> is hardly
-better than <a href="#TABLE-2">Table II</a> so that the value of variable interconnecting
-weights was not being fully realized. In a later experiment the number of elements
-was reduced to 12 elements and the same functions used. In this case
-the presence of extra interconnecting weights actually proved to be
-a hindrance! However a close examination of the incrementing process
-brought out the fact that the troublesome behavior was due to the
-greater chance of having only a few (often only one) elements do nearly
-all the incrementing. It is expected that the use of the additional
-refinements discussed herein will produce a considerable improvement
-in bringing out the full power of adaptation in multiple layers of a
-network.</p>
-
-<table id= "TABLE-2" border="0" cellspacing="0" summary="Table II" cellpadding="2" >
- <caption><big><b>TABLE II</b></big></caption>
- <thead><tr>
- <th class="tdc bb2" colspan="8">&nbsp;</th>
- </tr><tr>
- <th class="tdc br" colspan="2"><big>A</big></th>
- <th class="tdc br" colspan="2"><big>B</big></th>
- <th class="tdc br" colspan="2"><big>C</big></th>
- <th class="tdc" colspan="2"><big>D</big></th>
- </tr><tr>
- <th class="tdc bb">&emsp;r&emsp;</th> <th class="tdc bb br">e&nbsp;</th>
- <th class="tdc bb">&emsp;r&emsp;</th> <th class="tdc bb br">e</th>
- <th class="tdc bb">&emsp;r&emsp;</th> <th class="tdc bb br">e</th>
- <th class="tdc bb">&emsp;r&emsp;</th> <th class="tdc bb">e</th>
- </tr>
- </thead>
- <tbody><tr>
- <td class="tdc">7</td> <td class="tdc br">47&nbsp;</td>
- <td class="tdc">18&#8199;</td> <td class="tdc br">192&#8199;</td>
- <td class="tdc">8</td> <td class="tdc br">110&#8199;</td>
- <td class="tdc">4</td> <td class="tdc">48&nbsp;</td>
- </tr><tr>
- <td class="tdc">3</td> <td class="tdc br">40&nbsp;</td>
- <td class="tdc">7</td> <td class="tdc br">69&nbsp;</td>
- <td class="tdc">10&#8199;</td> <td class="tdc br">98&nbsp;</td>
- <td class="tdc">6</td> <td class="tdc">68&nbsp;</td>
- </tr><tr>
- <td class="tdc">4</td> <td class="tdc br">43&nbsp;</td>
- <td class="tdc">7</td> <td class="tdc br">82&nbsp;</td>
- <td class="tdc">4</td> <td class="tdc br">47&nbsp;</td>
- <td class="tdc">6</td> <td class="tdc">46&nbsp;</td>
- </tr><tr>
- <td class="tdc bt2" colspan="8">&nbsp;</td>
- </tr>
- </tbody>
-</table>
-
-<h3>FUTURE PROBLEMS</h3>
-
-<p>Aside from the previous question of deciding on network structure,
-there are several other questions that remain to be studied in learning
-networks.</p>
-
-<p>There is the question of requiring more than a single output from a
-network. If, say, two outputs are required for a given input, one
-+1 and the other -1, this runs into conflict with the incrementing
-process. Changes that aid one output may act against the other.
-<span class="pagenum" id="Page_45">[Pg 45]</span>
-Apparently the searching process depicted before with a varying bias
-must be considerably refined to find weight changes which act on
-all the outputs in the required way. This is far from an academic
-question because there will undoubtedly be numerous cases in which
-the greatest part of the input-output computation will have shared
-features for all output variables. Only at later levels do they need to
-be differentiated. Hence it is necessary to envision a single network
-producing multiple outputs rather than a separate network for each
-output variable if full efficiency is to be achieved.</p>
-
-<p>Another related question is that of using input variables that are
-either many-, or continuous-, valued rather than two-valued. No
-fundamental difficulties are discernible in this case, but the matter
-deserves some considerable study and experimentation.</p>
-
-<p>Another important question involves the use of a succession of inputs
-for producing an output. That is, it may be useful to allow time to
-enter into the network’s logical action, thus giving it a “dynamic” as
-well as “static” capability.</p>
-
-<hr class="chap x-ebookmaker-drop" />
-
-<div class="chapter">
-<p><span class="pagenum" id="Page_46">[Pg 46]</span></p>
-<h2 class="nobreak">Adaptive Detection of Unknown<br /> Binary Waveforms</h2>
-</div>
-
-<p class="f120"><b><span class="smcap">J. J. Spilker, Jr.</span></b></p>
-
-<p class="center space-below1"><i>Philco Western Development Laboratories<br />
-Palo Alto, California</i></p>
-
-<div class="blockquot">
-<p>This work was supported by the Philco WDL Independent Development
-Program. This paper, submitted after the Symposium, represents a more
-detailed presentation of some of the issues raised in the discussion
-sessions at the Symposium and hence, constitutes a worthwhile addition
-to the Proceedings.</p>
-</div>
-
-<h3>INTRODUCTION</h3>
-
-<p>One of the most important objectives in processing a stream of
-data is to determine and detect the presence of any invariant or
-quasi-invariant “features” in that data stream. These features are
-often initially unknown and must be “learned” from the observations.
-One of the simplest features of this form is a finite length signal
-which occurs repetitively, but not necessarily periodically with time,
-and has a waveshape that remains invariant or varies only slowly with
-time.</p>
-
-<p>In this discussion, we assume that the data stream has been
-pre-processed, perhaps by a detector or discriminator, so as to exhibit
-this type of repetitive (but unknown) waveshape or signal structure.
-The observed signal, however, is perturbed by additive noise or other
-disturbances. It is desired to separate the quasi-invariance of the
-data from the truly random environment. The repetitive waveform may
-represent, for example, the transmission of an unknown sonar or radar,
-a pulse-position modulated noise-like waveform, or a repeated code word.</p>
-
-<p>The problem of concern is to estimate the signal waveshape and to
-determine the time of each signal occurrence. We limit this discussion
-to the situation where only a single repetitive waveform is present
-and the signal sample values are binary. The observed waveform is
-assumed to be received at low signal-to-noise ratio so that a single
-observation of the signal (even if one knew precisely the arrival time)
-is not sufficient to provide a good estimate of the signal waveshape.
-The occurrence time of each signal is assumed to be random.
-<span class="pagenum" id="Page_47">[Pg 47]</span></p>
-
-<h3>THE ADAPTIVE DETECTION MACHINE</h3>
-
-<p>The purpose of this note is to describe very briefly a machine<a id="FNanchor_2" href="#Footnote_2" class="fnanchor">[2]</a>
-which has been implemented to recover the noise-perturbed binary waveform.
-A simplified block diagram of the machine is shown in <a href="#FIG_1D">Figure 1</a>.
-The experimental machine has been designed to operate on signals of 10³
-samples duration.</p>
-
-<p>Each analog input sample enters the machine at left and may either
-contain a signal sample plus noise or noise alone. In order to permit
-digital operation in the machine, the samples are quantized in a
-symmetrical three-level quantizer. The samples are then converted to
-vector form, <i>e.g.</i>, the previous 10³ samples form the vector
-components. A new input vector, <span class="bt2">Y⁽ⁱ⁾</span>, is formed at each sample instant.</p>
-
-<p>Define the signal sample values as s₁, s₂, ..., sₙ. The observed vector
-Y⁽ⁱ⁾ is then either (a) perfectly centered signal plus noise, (b)
-shifted signal plus noise, or (c) noise alone.</p>
-
-<table border="0" cellspacing="0" summary=" " cellpadding="2" >
- <tbody><tr>
- <td class="tdc">&nbsp;</td>
- <td class="tdc" rowspan="3"><img src="images/cbl-3.jpg" alt="" width="16" height="57" /></td>
- <td class="tdl">(s₁, s₂, ..., sₙ) + (n₁, n₂, ..., nₙ)</td>
- <td class="tdr">(a)</td>
- </tr><tr>
- <td class="tdc">(Y⁽ⁱ⁾)ᵗ&nbsp;=&nbsp;</td>
- <td class="tdl">(0, ..., s₁, s₂, ..., sₙ₋ⱼ) + (n₁, n₂, ..., nₙ)</td>
- <td class="tdr">&nbsp;&nbsp;(b)</td>
- </tr><tr>
- <td class="tdc">&nbsp;</td>
- <td class="tdl">(0 ... 0) + (n₁, n₂, ..., nₙ)</td>
- <td class="tdr">(c)</td>
- </tr>
- </tbody>
-</table>
-
-<p>At each sample instant, two measurements are made on the input
-vector, an energy measurement <big>‖Y⁽ⁱ⁾‖²</big> and a polarity coincidence
-cross-correlation with the present estimate of the signal vector stored
-in memory. If the weighted sum of the energy and cross-correlation
-measurements exceeds the present threshold value <big>Γᵢ</big>, the input vector
-is accepted as containing the signal (properly shifted in time), and the
-input vector is added to the memory. The adaptive memory has <big>2<sup>Q</sup></big>
-levels, <big>2<sup>Q-1</sup></big> positive levels, 1 zero level and <big>2<sup>Q-1</sup>-1</big>
-negative levels. New contributions are made to the memory by normal vector
-addition except that saturation occurs when a component value is at the
-maximum or minimum level.</p>
-
-<p>The acceptance or rejection of a given input vector is based on a
-hypersphere decision boundary. The input vector is accepted if the
-weighted sum <big>γᵢ</big> exceeds the threshold <big>Γᵢ</big></p>
-
-<p class="f150">γᵢ&nbsp;=&nbsp;Y⁽ⁱ⁾∙M⁽ⁱ⁾&nbsp;+&nbsp;α‖Y⁽ⁱ⁾‖²&nbsp;⩾&nbsp;Γᵢ.</p>
-
-<p class="space-above2"><span class="pagenum" id="Page_48">[Pg 48]</span></p>
-<div class="figcenter">
- <img id="FIG_1D" src="images/i_055.jpg" alt="" width="600" height="252" />
- <p class="f120 space-below2">Figure 1—Block diagram of the adaptive
- binary waveform detector</p>
-</div>
-
-<p class="space-below1"><span class="pagenum" id="Page_49">[Pg 49]</span>
-Geometrically, we see that the input vector is accepted if it falls on
-or outside of a hypersphere centered at</p>
-
-<table class="fontsize_130" border="0" cellspacing="0" summary=" " cellpadding="0" >
- <tbody><tr>
- <td class="tdc">&nbsp;</td>
- <td class="tdc">-&nbsp;<span class="bt2">M⁽ⁱ⁾</span></td>
- </tr><tr>
- <td class="tdc"><span class="bt2">C⁽ⁱ⁾</span> =</td>
- <td class="tdc">&nbsp;&mdash;&mdash;</td>
- </tr><tr>
- <td class="tdc">&nbsp;</td>
- <td class="tdc">2α</td>
- </tr>
- </tbody>
-</table>
-<p class="neg-indent">having radius squared</p>
-
-<table class="fontsize_130" border="0" cellspacing="0" summary=" " cellpadding="2" >
- <tbody><tr>
- <td class="tdc">&nbsp;</td>
- <td class="tdc">Γ⁽ⁱ⁾</td>
- <td class="tdc">&nbsp;</td>
- <td class="tdc">‖M⁽ⁱ⁾‖²</td>
- </tr><tr>
- <td class="tdc">[r⁽ⁱ⁾]² =</td>
- <td class="tdc">&nbsp;——</td>
- <td class="tdc">&nbsp;+&nbsp;</td>
- <td class="tdc">———— .</td>
- </tr><tr>
- <td class="tdc">&nbsp;</td>
- <td class="tdc">α</td>
- <td class="tdc">&nbsp;</td>
- <td class="tdc">(2α)²</td>
- </tr>
- </tbody>
-</table>
-
-<p>Both the center and radius of this hypersphere change as the machine
-adapts. The performance and optimality of hypersphere-type decision
-boundaries have been <i>discussed in related work</i> by Glaser<a id="FNanchor_3" href="#Footnote_3" class="fnanchor">[3]</a>
-and Cooper.<a id="FNanchor_4" href="#Footnote_4" class="fnanchor">[4]</a></p>
-
-<p>The threshold value, <b>Γᵢ</b>, is adapted so that it increases if the
-memory becomes a better replica of the signal with the result that <big><b>γᵢ</b></big>
-increases. On the other hand, if the memory is a poor replica of the
-signal (for example, if it contains noise alone), it is necessary that
-the threshold decay with time to the point where additional acceptances
-can modify the memory structure.</p>
-
-<p>The experimental machine is entirely digital in operation and, as
-stated above, is capable of recovering waveforms of up to 10³ samples
-in duration. In a typical experiment, one might attempt to recover
-an unknown noise-perturbed, pseudo-random waveform of up to 10³ bits
-duration which occurs at random intervals. If no information is
-available as to the signal waveshape, the adaptive memory is blank at
-the start of the experiment.</p>
-
-<p>In order to illustrate the operation of the machine most clearly, let
-us consider a repetitive binary waveform which is composed of 10³ bits
-of alternate “zeros” and “ones.” A portion of this waveform is shown in
-<a href="#FIG_2D">Figure 2a</a>. The waveform actually observed is a noise-perturbed
-version of this waveform shown in <a href="#FIG_2D">Figure 2b</a> at-6 db signal-to-noise
-ratio. The exact sign of each of the signal bits obviously could not be accurately
-determined by direct observation of <a href="#FIG_2D">Figure 2b</a>.</p>
-
-<div class="figcenter">
- <img id="FIG_2D" src="images/i_056a.jpg" alt="" width="500" height="150" />
- <p class="f120 space-below2">(a) Binary signal</p>
- <img src="images/i_056b.jpg" alt="" width="500" height="155" />
- <p class="f120 space-below2">(b) Binary signal plus noise</p>
- <p class="f120 space-below2">Figure 2—Binary signal with additive noise at-6 db SNR</p>
-</div>
-<p><span class="pagenum" id="Page_50">[Pg 50]</span></p>
-<div class="figcontainer">
- <div class="figsub">
- <img id="FIG_3D" src="images/i_057a.jpg" alt="" width="250" height="166" />
- <p class="f120">(a)</p>
- </div>
- <div class="figsub">
- <img src="images/i_057b.jpg" alt="" width="250" height="166" />
- <p class="f120">(b)</p>
- </div>
-</div>
-
-<div class="figcontainer">
- <div class="figsub">
- <img src="images/i_057c.jpg" alt="" width="251" height="167" />
- <p class="f120">(c)</p>
- </div>
- <div class="figsub">
- <img src="images/i_057d.jpg" alt="" width="251" height="167" />
- <p class="f120">(d)</p>
- </div>
-</div>
-<div class="figcontainer">
- <img src="images/i_057e.jpg" alt="" width="500" height="162" />
- <p class="f120">(e)</p>
- <div class="blockquot">
- <p class="f120">Figure 3—Adaption of the memory at-6 db SNR: (a) Blank initial
- memory; (b) Memory after first dump; (c) Memory after 12 dumps; (d)
- Memory after 40 dumps; (e) Perfect “checkerboard” memory for comparison</p>
- </div>
-</div>
-
-<p>As the machine memory adapts to this noisy input signal, it progresses
-as shown in <a href="#FIG_3D">Figure 3</a>. The sign of 10<sup>3</sup> memory components
-are displayed in a raster pattern in this figure. <a href="#FIG_3D">Figure 3a</a> shows the
-memory in its blank initial state at the start of the adaption process.
-<a href="#FIG_3D">Figure 3b</a> shows the memory after the first adaption of the
-memory. This first “dump” occurred after the threshold had decayed to the point
-where an energy measurement produced an acceptance decision. <a href="#FIG_3D">Figure 3c</a>
-<span class="pagenum" id="Page_51">[Pg 51]</span>
-and 3d show the memory after 12 and 40 adaptions, respectively. These
-dumps, of course, are based on both energy and cross-correlation
-measurements. As can be seen, the adapted memory after 40 dumps is
-already quite close to the perfect memory shown by the “checkerboard”
-pattern of <a href="#FIG_3D">Figure 3c</a>.</p>
-
-<p>The detailed analysis of the performance of this type of machine
-vs. signal-to-noise ratio, average signal repetition rate, signal
-duration, and machine parameters is extremely complex. Therefore, it
-is not appropriate here to detail the results of the analytical and
-experimental work on the performance of this machine. However, several
-conclusions of a general nature can be stated.</p>
-
-<div class="blockquot">
-<p>(a) Because the machine memory is always adapting, there is
-a relatively high penalty for “false alarms.” False alarms can
-destroy a perfect memory. Hence, the threshold level needs to be set
-appropriately high for the memory adaption. If one wishes to detect
-signal occurrences with more tolerance to false alarms, a separate
-comparator and threshold level should be used.</p>
-
-<p>(b) The present machine structure, which allows for slowly varying
-changes in the signal waveshape, exhibits a marked threshold effect
-in steady-state performance at an input signal-to-noise ratio (peak
-signal power-to-average noise power ratio) of about -12 db. Below
-this signal level, the time required for convergence increases very
-rapidly with decreasing signal level. At higher SNR, convergence to
-noise-like signals, having good auto-correlation properties, occurs at
-a satisfactory rate.</p>
-</div>
-
-<p>A more detailed discussion of performance has been published in the
-report cited in footnote reference 1.</p>
-
-<hr class="chap x-ebookmaker-drop" />
-<div class="chapter">
-<p><span class="pagenum" id="Page_52">[Pg 52]</span></p>
-
-<h2 class="nobreak">Conceptual Design of Self-Organizing Machines</h2>
-</div>
-
-<p class="f120"><b><span class="smcap">P. A. Kleyn</span></b></p>
-
-<p class="center space-below1"><i>Northrop Nortronics</i><br />
-<i>Systems Support Department</i><br /><i>Anaheim, California</i></p>
-
-<div class="blockquot">
-<p>Self-organization is defined and several examples which motivate
-this definition are presented. The significance of this definition
-is explored by comparison with the metrization problem discussed
-in the companion paper <a href="#REF_E_1">(1)</a> and it is seen that self-organization
-requires decomposing the space representing the environment. In the
-absence of a priori knowledge of the environment, the self-organizing
-machine must resort to a sequence of projections on unit spheres to
-effect this decomposition. Such a sequence of projections can be
-provided by repeated use of a nilpotent projection operator (NPO). An
-analog computer mechanization of one such NPO is discussed and the
-signal processing behavior of the NPO is presented in detail using
-the Euclidean geometrical representation of the metrizable topology
-provided in the companion paper. Self-organizing systems using multiple
-NPO’s are discussed and current areas of research are identified.</p>
-</div>
-
-<h3>INTRODUCTION</h3>
-
-<p>Unlike the companion paper which considers certain questions in
-depth, this paper presents a survey of the scope of our work in
-self-organizing systems and is not intended to be profound.</p>
-
-<p>The approach we have followed may be called phenomenological (<a href="#FIG_1E">Figure 1</a>).
-That is, the desired behavior (self-organization) was defined,
-represented mathematically, and a mechanism(s) required to yield the
-postulated behavior was synthesized using mathematical techniques. One
-advantage of this approach is that it avoids assumptions of uniqueness
-of the mechanism. Another advantage is that the desired behavior, which
-is after all the principal objective, is taken as invariant. An obvious
-disadvantage is the requirement for the aforementioned synthesis
-technique; fortunately in our case a sufficiently general technique had
-been developed by the author of the companion paper.</p>
-
-<p>From the foregoing and from the definition of self-organization we
-employ (<a href="#CONC_MOD">see conceptual model</a>), it would appear that our research does
-<span class="pagenum" id="Page_53">[Pg 53]</span>
-not fit comfortably within any of the well publicized approaches to
-self-organization <a href="#REF_E_2">(2)</a>. Philosophically, we lean toward viewpoints
-expressed by <a href="#REF_E_3">Ashby (3)</a>, <a href="#REF_E_4">(4)</a>,
-<a href="#REF_E_5">Hawkins (5)</a>, and <a href="#REF_E_6">Mesarovic (6)</a>
-but with certain reservations. We have avoided the neural net approach partly
-because it is receiving considerable attention and also because the
-brain mechanism need not be the unique way to produce the desired
-behavior.</p>
-
-<div class="figcenter">
- <img id="FIG_1E" src="images/i_060.jpg" alt="" width="600" height="330" />
- <p class="f120 space-below2">Figure 1—Approach used in Nortronics research
- on self-organizing systems</p>
-</div>
-
-<p class="space-below2">Nor have we followed the probability
-computer or statistical decision theory approach exemplified by
-<a href="#REF_E_7">Braverman (7)</a> because these usually require some sort of preassigned
-coordinate system <a href="#REF_E_8">(8)</a>. Neither will the reader find much indication
-of formal logic <a href="#REF_E_9">(9)</a> or heuristic <a href="#REF_E_10">(10)</a>
-programming. Instead, we view a self-organizing system more as a mirror
-whose appearance reflects the environment rather than its own intrinsic
-nature. With this viewpoint, a self-organizing system appears very
-flexible because it possesses few internal constraints which would tend
-to distort the reflection of the environment and hinder its ability to adapt.</p>
-
-<h3 id="CONC_MOD">CONCEPTUAL MODEL</h3>
-
-<p class="f120"><b>Definition</b></p>
-
-<p>A system is said to be self-organizing if, after observing the input
-and output of an unknown phenomenon (transfer relation), the system
-organizes itself into a simulation of the unknown phenomenon.</p>
-
-<p>Implicit in this definition is the requirement that the self-organizing
-machine (SOM) not possess a preassigned coordinate system. In fact it
-is just this ability to acquire that coordinate system implicit in the
-input-output spaces which define the phenomenon that we designate as
-<span class="pagenum" id="Page_54">[Pg 54]</span>
-self-organization. Thus any a priori information programmed into the
-SOM by means of, for example, stored or wired programs, constrains
-the SOM and limits its ability to adapt. We do not mean to suggest
-that such preprogramming is not useful or desirable; merely that it is
-inconsistent with the requirement for self-organization. As shown in
-<a href="#FIG_2E">Figure 2</a>, it is the given portion of the environment which
-the SOM is to simulate, which via the defining end spaces, furnishes the SOM with
-all the data it needs to construct the coordinate system intrinsic to
-those spaces.</p>
-
-<p>The motivation for requiring the ability to simulate as a feature of
-self-organization stems from the following examples.</p>
-
-<p>Consider the operation of driving an automobile. <a href="#FIG_3E">Figure 3</a>
-depicts the relation characterized by a set of inputs; steering, throttle, brakes,
-transmission, and a set of outputs; the trajectory. Operation of the
-automobile requires a device (SOM) which for a desired trajectory can
-furnish those inputs which realize the desired trajectory. In order to
-provide the proper inputs to the automobile, the SOM must contain a
-simulation of <big><b>⨍⁻¹(x)</b></big>.</p>
-
-<div class="figcenter">
- <img id="FIG_2E" src="images/i_061a.jpg" alt="" width="600" height="131" />
- <p class="f120">Figure 2—Simulation of (a portion of) the environment</p>
- <img id="FIG_3E" src="images/i_061b.jpg" alt="" width="600" height="173" />
- <p class="f120">Figure 3—Simulation of a relation</p>
-</div>
-
-<p>Since <big><b>⨍(x)</b></big> is completely defined in terms of the inputs
-and the resulting trajectories, exposure to them provide the SOM with all the
-information necessary to simulate <big><b>⨍⁻¹(x)</b></big>. And if the SOM
-possesses internal processes which cause rearrangement of the input-output
-relation of the SOM to correspond to <big><b>⨍⁻¹(x)</b></big> in accordance with
-the observed data, the SOM can operate an automobile. It is this internal
-change which is implied by the term “self-organizing,” but note that the
-<span class="pagenum" id="Page_55">[Pg 55]</span>
-instructions which specify the desired organization have their source
-in the environment.</p>
-
-<p>As a second example consider adaptation to the environment. Adapt
-(from Webster) means: “to change (oneself) so that one’s behavior,
-attitudes, <i>etc.</i>, will conform to new or changed circumstances.
-Adaptation in biology means a change in structure, function or form
-that produces better adjustment to the environment.” These statements
-suggest a simulation because adjustment to the environment implies
-survival by exposing the organism to the beneficial rather than the
-inimical effects of the environment. If we represent the environment
-(or portion thereof) as a relation as shown in <a href="#FIG_2E">Figure 2</a>,
-we note that the ability to predict what effect a given disturbance will have
-is due to a simulation of the cause-effect relation which characterizes the
-environment.</p>
-
-<p>It would be a mistake to infer from these examples that simulation
-preserves the appearance of the causes and effects which characterize
-a relation. We clarify this situation by examining a relation and its
-simulation.</p>
-
-<p>Consider the relation between two mothers and their sons as pictured
-in <a href="#FIG_4E">Figure 4</a>. Observe that if symbols (points) are substituted
-for the actual physical objects (mothers and sons), the relation is not altered
-in any way. This is what we mean by simulation and this is how a SOM
-simulates. It is not even necessary that the objects, used to display
-the relation, be defined; <i>i.e.</i>, these objects may be primitive.
-(If this were not so, no mathematical or physical theory could model
-the environment.) The main prerequisite is sufficient resolution to
-distinguish the objects from each other.</p>
-
-<div class="figcenter">
- <img id="FIG_4E" src="images/i_062.jpg" alt="" width="600" height="385" />
- <p class="f120">Figure 4—A relation of objects—displayed and simulated</p>
-</div>
-<p class="space-below2"><span class="pagenum" id="Page_56">[Pg 56]</span></p>
-
-<h3>MATHEMATICAL MODEL</h3>
-
-<p>The mathematical model must represent both the environment and the SOM
-and for reasons given in the companion paper each is represented as a
-metrizable topology. For uniqueness we factor each space into equal
-parts and represent the environment as the channel</p>
-
-<p class="f120">W ⟶ X.&emsp;<small>(Ref. 10a)</small></p>
-
-<p>Consider now the SOM to be represented by the cascaded channels</p>
-
-<p class="f120">X ⟶ Y ⟶ Z</p>
-
-<p class="no-indent">where <big>X&nbsp;⟶&nbsp;Y</big> is a variable which
-represents the reorganization of the SOM existing input-output relation
-represented by <big>Y&nbsp;⟶&nbsp;Z</big>.</p>
-
-<p>The solution of the three channels-in-cascade problem</p>
-
-<p class="f120">W ⟶ X ⟶ Y ⟶ Z,</p>
-
-<p class="no-indent">where <big>p(W) (11), p(X), p(X|W), p(Y), p(Z),
-p(Z|Y)</big> are fixed, yields that middle channel <big>p₀(Y|X)</big>,
-from a set of permissible middle channels <big>{p(Y|X)}</big>, which
-maximizes <big>R(Z,W)</big>.</p>
-
-<p>Then the resulting middle channel describes that reorganization of
-the SOM which yields the optimum simulation of <big>W ⟶ X</big> by the
-SOM, within the constraints upon <big>Ch(Z,Y)</big>.</p>
-
-<p>The solution (the middle channel) depends of course on the particular
-end channels. Obviously the algorithm which is used to find the
-solution does not. It follows that if some physical process were
-constrained to carrying out the steps specified by the algorithm,
-said process would be capable of simulation and would exhibit
-self-organization.</p>
-
-<p>Although the formal solution to the three-channels-in-cascade problem
-is not complete, the solution is sufficiently well characterized to
-permit proceeding with a mechanization of the algorithm. A considerable
-portion of the solution is concerned with the decomposition and
-metrization of channels and it is upon this feature that we now focus
-attention.</p>
-
-<p>As suggested in the companion paper, if the dimensionality of the
-spaces is greater than one, the SOM has only one method available (12).
-Consider the decomposition of a space without, for the moment, making
-the distinction between input and output.</p>
-
-<p><a href="#FIG_5E">Figure 5</a> depicts objects represented by a (perhaps multidimensional)
-“cloud” of points. In the absence of a preassigned coordinate system,
-<span class="pagenum" id="Page_57">[Pg 57]</span>
-the SOM computes the center of gravity of the cloud (which can be
-done in any coordinate system) and describes the points in terms of
-the distance from this center of gravity; or, which is the same, as
-concentric spheres with origin at the center of gravity.</p>
-
-<div class="figcenter">
- <img id="FIG_5E" src="images/i_064.jpg" alt="" width="600" height="337" />
- <p class="f120">Figure 5—Nilpotent decomposition of a three-dimensional space</p>
-</div>
-
-<p>The direction of particular point cannot be specified for there is no
-reference radius vector. Since the SOM wants to end up with a cartesian
-coordinate system, it must transform the sphere (a two-dimensional
-surface) into a plane (a two-dimensional surface). Unfortunately, a
-sphere is not homeomorphic to a plane; thus the SOM has to decompose
-the sphere into a cartesian product of a hemisphere <a href="#REF_E_12A">(12a)</a>
-and a denumerable group. The SOM then can transform the hemisphere into a
-plane. The points projected onto the plane constitute a space of the
-same character as the one with which the SOM started. Thus, it can
-repeat all operations on the plane (a space of one less dimension) by
-finding the center of gravity and the circle upon which the desired
-point is situated. The circle is similarly decomposed into a line times
-a denumerable group. By repeating this operation as many times as the
-space has dimensions, the SOM eventually arrives at a single point and
-has obtained in the process a description of the space. Since this
-procedure can be carried on by the repeated use of one operator, this
-operator is nilpotent and to reflect this fact as well as the use of a
-projection, we have named this a nilpotent projection operator or NPO
-for short.</p>
-
-<h3>MECHANIZATION OF THE NPO</h3>
-
-<p>Analog computer elements were used to simulate one NPO which
-was tested <span class="pagenum" id="Page_58">[Pg 58]</span> in
-the experimental configuration shown in <a href="#FIG_6E">Figure 6</a>. The NPO
-operates upon a channel which is artificially generated from the two noise
-generators <big>i₁</big> and <big>i₂</big> and the signal generator
-<big>i₀</big> (<big>i₀</big> may also be a noise generator). The
-NPO accepts the inputs labelled <big>X₁</big> and <big>X₂</big> and
-provides the three outputs <big>Ξ₁, Ξ₂</big>, and <big>γ.&emsp;X₁</big> is
-the linear combination of the outputs of generators <big>i₁</big> and
-<big>i₀</big>, similarly <big>X₂</big> is obtained from <big>i₂</big>
-and <big>i₀</big>.</p>
-
-<div class="figcenter">
- <img id="FIG_6E" src="images/i_065.jpg" alt="" width="600" height="288" />
- <p class="f120 space-below2">Figure 6—Experimental test configuration for the
- simulation of an NPO</p>
-</div>
-
-<p>Obviously, <big>i₀</big> is an important parameter since it
-represents the memory relating the spaces <big>X₁</big> and
-<big>X₂</big>. <big>Ξ₁</big> has the property that the magnitude of its
-projection on i₀ is a maximum while <big>Ξ₂</big> to the opposite has a
-zero projection on <big>i₀</big>. <big>γ</big> is the detected version
-of the eigenvalue of <big>Ch(X₂,X₁)</big>.</p>
-
-<p>In the companion paper it was shown how one can provide a Euclidean
-geometrical representation of the NPO. This representation is shown in
-<a href="#FIG_7E">Figure 7</a> which shows the vectors <big>i₀, i₁, i₂, X₁, X₂, Ξ₁, Ξ₂,</big>
-and the angles <big>Θ₁, Θ₂,</big> and <big>γ</big>. The length of a
-vector is given by</p>
-
-<p class="f120">|X| = κₓ(2πε)⁻¹ᐟ² ∈ H(X)</p>
-
-<p class="no-indent">and the angle between two vectors by</p>
-
-<p class="f120">|Θ(X₁,X₂)|-sin⁻¹ ∈ -R(X₁,X₂).</p>
-
-<p>The three vectors <big>i₀, i₁, i₂</big> provide an orthogonal coordinate system
-because the corresponding signals are random, <i>i.e.</i>,</p>
-
-<table class="fontsize_130" border="0" cellspacing="0" summary=" " cellpadding="0" >
- <tbody><tr>
- <td class="tdl">&nbsp;</td>
- <td class="tdc">κ</td>
- <td class="tdc">&nbsp;</td>
- </tr><tr>
- <td class="tdl">R(i₀,i₁,i₂)</td>
- <td class="tdc">&nbsp;≡&nbsp;</td>
- <td class="tdc">0.</td>
- </tr>
- </tbody>
-</table>
-
-<p>As external observers we have a prior knowledge of this coordinate
-<span class="pagenum" id="Page_59">[Pg 59]</span> system; however,
-the NPO is given only the vectors <big>X₁</big> and <big>X₂</big> in
-the &nbsp;<big>i₀ ⨉ i₁</big> &nbsp;and &nbsp;<big>i₀ ⨉ i₂</big> planes.
-The NPO can reconstruct the entire geometry but the actual output <big>Ξ</big> obviously
-is constrained to lie in the plane of the input vector <big>X</big>. The
-following formulas are typical of the relations present.</p>
-
-<table class="fontsize_130" border="0" cellspacing="0" summary=" " cellpadding="0" >
- <tbody><tr>
- <td class="tdl">&nbsp;</td>
- <td class="tdc">&nbsp;</td>
- <td class="tdl">&nbsp;|Ξ₁|</td>
- </tr><tr>
- <td class="tdl">tan β</td>
- <td class="tdc">&nbsp;=&nbsp;</td>
- <td class="tdl">——</td>
- </tr><tr>
- <td class="tdl">&nbsp;</td>
- <td class="tdc">&nbsp;</td>
- <td class="tdl">&nbsp;|Ξ₂|</td>
- </tr><tr>
- <td class="tdc" colspan="3">&nbsp;</td>
- </tr><tr>
- <td class="tdl">cos Θ</td>
- <td class="tdc">=</td>
- <td class="tdc">cos 2β csc 2γ</td>
- </tr><tr>
- <td class="tdc" colspan="3">&nbsp;</td>
- </tr>
- </tbody>
-</table>
-
-<table class="fontsize_130" border="0" cellspacing="0" summary=" " cellpadding="0" >
- <tbody><tr>
- <td class="tdl">&nbsp;</td>
- <td class="tdc">&nbsp;</td>
- <td class="tdc">&nbsp;</td>
- <td class="tdc">cos 2β</td>
- </tr><tr>
- <td class="tdl">cos 2Θ₁</td>
- <td class="tdc">&nbsp;=&nbsp;</td>
- <td class="tdc">-1 + 2&nbsp;</td>
- <td class="tdc">———</td>
- </tr><tr>
- <td class="tdl">&nbsp;</td>
- <td class="tdc">&nbsp;</td>
- <td class="tdc">&nbsp;</td>
- <td class="tdc">1-cos 2γ</td>
- </tr><tr>
- <td class="tdc" colspan="4">&nbsp;</td>
- </tr>
- </tbody>
-</table>
-
-<p class="center fontsize_130">cos Θ = cos Θ₁ cos Θ₂.</p>
-
-<div class="figcenter">
- <img id="FIG_7E" src="images/i_066.jpg" alt="" width="550" height="520" />
- <p class="f120 space-below2">Figure 7—Geometry of the NPO</p>
-</div>
-<p><span class="pagenum" id="Page_60">[Pg 60]</span></p>
-
-<div class="figcenter">
- <img id="FIG_8E" src="images/i_067a.jpg" alt="" width="400" height="489" />
- <p class="f120 space-below2">Figure 8—NPO run number 5</p>
- <img id="FIG_9E" src="images/i_067b.jpg" alt="" width="400" height="541" />
- <p class="f120 space-below2">Figure 9—NPO run number 6</p>
-</div>
-<p><span class="pagenum" id="Page_61">[Pg 61]</span></p>
-
-<p>We have obtained a complete description of the NPO which involves
-74 formulas. These treat the noise in the various outputs, invariances
-of the NPO and other interesting features. A presentation of these
-would be outside of the scope of this paper and would tend to obscure
-the main features of the NPO. Thus, we show here only a typical
-sample of the computer simulation, <a href="#FIG_8E">Figure 8</a>
-and <a href="#FIG_9E">Figure 9</a>. Conditions for these runs are shown
-in <a href="#TABLE_1">Table I</a>. Run No. 6 duplicates run No. 5 except for
-the fact that <big>i₁</big> and <big>i₂</big> were disabled in run No. 6.</p>
-
-<p>Observe that all our descriptions of the NPO and the space it is to
-decompose have been time invariant while the signals shown in the
-simulation are presented as functions of time. The conversion may be
-effected as follows: Given a measurable (single-valued) function</p>
-
-<p class="f120">x = x(t)t ∊ T</p>
-
-<p class="no-indent">where</p>
-
-<p class="f120">μ(T) &gt; 0</p>
-
-<p class="no-indent">we define the space</p>
-
-<p class="f120">X = <big>{</big>x = x(t) ∍ t ∊ T<big>}</big></p>
-
-<p class="no-indent">and a probability distribution</p>
-
-<table class="fontsize_130" border="0" cellspacing="0" summary=" " cellpadding="0" >
- <tbody><tr>
- <td class="tdl">&nbsp;</td>
- <td class="tdc">μ(x⁻¹(X′))</td>
- <td class="tdc">&nbsp;</td>
- </tr><tr>
- <td class="tdl">P(X′) =&nbsp;</td>
- <td class="tdl">&nbsp;————</td>
- <td class="tdl">&nbsp;X′ open ⊂ X</td>
- </tr><tr>
- <td class="tdl">&nbsp;</td>
- <td class="tdc">μ(T)</td>
- <td class="tdc">&nbsp;</td>
- </tr><tr>
- <td class="tdc" colspan="3">&nbsp;</td>
- </tr>
- </tbody>
-</table>
-<p class="no-indent">on that space.</p>
-
-<table id="TABLE_1" border="0" cellspacing="0" summary="Table I" cellpadding="0" rules="cols" >
- <caption><big><b>TABLE I</b></big></caption>
- <thead><tr>
- <th class="tdc bb" colspan="8">Legend for Traces of Figures 8 and 9</th>
- </tr><tr>
- <th class="tdc bb">Trace Number</th>
- <th class="tdc bb">1</th>
- <th class="tdc bb">2</th>
- <th class="tdc bb">3</th>
- <th class="tdc bb">4</th>
- <th class="tdc bb">5</th>
- <th class="tdc bb">6</th>
- <th class="tdc bb">7</th>
- </tr><tr>
- <th class="tdc bb2">Symbol</th>
- <th class="tdc bb2">X₂</th>
- <th class="tdc bb2">X₁</th>
- <th class="tdc bb2">γ</th>
- <th class="tdc bb2">β</th>
- <th class="tdc bb2">i</th>
- <th class="tdc bb2">&nbsp;dξ₂/dτ&nbsp;</th>
- <th class="tdc bb2">&nbsp;dξ₁/dτ</th>
- </tr>
- </thead>
- <tbody><tr>
- <td class="tdl" colspan="8">run No. 5</td>
- </tr><tr>
- <td class="tdl">signal</td> <td class="tdc">&nbsp;7½ Vrms&nbsp;</td>
- <td class="tdc">&nbsp;7½ Vrms&nbsp;</td> <td class="tdc">π ptop</td>
- <td class="tdc">&nbsp;</td> <td class="tdc">&nbsp;35.6 m cps&nbsp;</td>
- <td class="tdc">&nbsp;</td> <td class="tdc">&nbsp;</td>
- </tr><tr>
- <td class="tdl">noise</td> <td class="tdc">16 Vrms</td>
- <td class="tdc">15 Vrms</td>
- <td class="tdc">&nbsp;π/9 ptop<a id="FNanchor_5" href="#Footnote_5" class="fnanchor">[5]</a>&nbsp;</td>
- <td class="tdc">&nbsp;</td> <td class="tdc">sine wave</td>
- <td class="tdc">&nbsp;</td> <td class="tdc">&nbsp;</td>
- </tr><tr>
- <td class="tdl">DC</td> <td class="tdc">0</td>
- <td class="tdc">0</td> <td class="tdc">&nbsp;</td>
- <td class="tdc">&nbsp;</td> <td class="tdc">&nbsp;</td>
- <td class="tdc">&nbsp;</td> <td class="tdc">&nbsp;</td>
- </tr><tr>
- <td class="tdl">power s/n</td> <td class="tdc">1/4</td>
- <td class="tdc">1/4</td> <td class="tdc">81/1</td>
- <td class="tdc">&nbsp;</td> <td class="tdc">&nbsp;</td>
- <td class="tdc">0</td>
- <td class="tdc">1/2<a id="FNanchor_6" href="#Footnote_6" class="fnanchor">[6]</a></td>
- </tr><tr>
- <td class="tdl bb">terminal value</td> <td class="tdc bb">&nbsp;</td>
- <td class="tdc bb">&nbsp;</td> <td class="tdc bb">π/4</td>
- <td class="tdc bb">&nbsp;&nbsp;π/4&nbsp;&nbsp;</td> <td class="tdc bb">&nbsp;</td>
- <td class="tdc bb">&nbsp;</td> <td class="tdc bb">&nbsp;</td>
- </tr><tr>
- <td class="tdl" colspan="8">run No. 6</td>
- </tr><tr>
- <td class="tdl">signal</td> <td class="tdc">7½ Vrms</td>
- <td class="tdc">7½ Vrms</td> <td class="tdc">π ptop</td>
- <td class="tdc">&nbsp;</td> <td class="tdc">35.6 m cps</td>
- <td class="tdc">&nbsp;</td> <td class="tdc">&nbsp;</td>
- </tr><tr>
- <td class="tdl">noise</td> <td class="tdc">0</td>
- <td class="tdc">0</td>
- <td class="tdc">0<a id="FNanchor_7" href="#Footnote_7" class="fnanchor">[7]</a></td>
- <td class="tdc">&nbsp;</td> <td class="tdc">sine wave</td>
- <td class="tdc">&nbsp;</td> <td class="tdc">&nbsp;</td>
- </tr><tr>
- <td class="tdl">DC</td> <td class="tdc">-30V</td>
- <td class="tdc">0</td> <td class="tdc">&nbsp;</td>
- <td class="tdc">&nbsp;</td> <td class="tdc">&nbsp;</td>
- <td class="tdc">&nbsp;</td> <td class="tdc">&nbsp;</td>
- </tr><tr>
- <td class="tdl">power s/n</td> <td class="tdc">∞</td>
- <td class="tdc">∞</td> <td class="tdc">∞</td>
- <td class="tdc">&nbsp;</td> <td class="tdc">&nbsp;</td>
- <td class="tdc">0</td> <td class="tdc">∞</td>
- </tr><tr>
- <td class="tdl">terminal value</td> <td class="tdc">&nbsp;</td>
- <td class="tdc">&nbsp;</td> <td class="tdc">π/4</td>
- <td class="tdc">π/4</td> <td class="tdc">&nbsp;</td>
- <td class="tdc">&nbsp;</td> <td class="tdc">&nbsp;</td>
- </tr><tr>
- <td class="tdc bt2" colspan="8">&nbsp;</td>
- </tr>
- </tbody>
-</table>
-<p><span class="pagenum" id="Page_62">[Pg 62]</span></p>
-
-<p>Then <big>(X,p(X))</big> is a stochastic space in our usual sense
-and <big>x(T)</big> is a stochastic variable. Two immediate consequences are:</p>
-
-<p><big>P(X)</big> is stationary <big>(P(X)</big> is not a function of
-&nbsp; <big>t ∊ T)</big>, and no question of ergodicity arises.</p>
-
-<h3>NETWORKS OF NPO’S</h3>
-
-<p>A network of NPO’s may constitute anything from a SOM to a
-preprogrammed detector, depending upon the relative amount of
-preprogramming included. Two methods of preprogramming are: (1) Feeding
-a signal out of a permanent storage into some of the inputs of the
-network of NPO’s. This a priori copy need not be perfect, because the
-SOM will measure the angles <big>Θᵢ</big> anyhow. (2) Feedback, which, after all,
-is just a way of taking advantage of the storage inherent in any delay
-line. (We implicitly assume that any reasonable physical realization
-of an NPO will include a delay <big>T</big> between the <big>x input</big>
-and the <big>ξ output</big> which is not less than perhaps 10⁻¹ times the
-time constant of the internal feedback loop in the <big>γ</big> computation.)</p>
-
-<p>Simulation of channels that possess a discrete component requires
-feedback path(s) to generate the required free products of the finitely
-generated groups. Then, such a SOM converges to a maximal subgroup of
-the group describing the symmetry of the signal that is a free product
-available to this SOM.</p>
-
-<p>Because a single NPO with <big>1 ≤ n₀ ≤ K₀</big> is isomorphic (provides the same
-input to output mapping) to a suitable network of NPO’s with <big>n₀ = 1</big>, it
-suffices to study only networks of NPO’s with <big>n₀ = 1</big>.</p>
-
-<p><a href="#FIG_10E">Figure 10</a> is largely self-explanatory. Item a is our schematic
-symbol for a single NPO with <big>n₀ = 1</big>. Items b, d (including larger feedback
-loops), and f are typical of artificial intelligence networks. Item c
-is employed to effect the level changing required in order to apply the
-three channels in cascade algorithm to the solution of one-dimensional
-coding problems. Observe that items c and e are the only configurations
-requiring the <big>γ</big> output. Item d may be used as a limiter by making <big>T⁻¹</big>
-high compared to the highest frequency present in the signal. Observe
-that item e is the only application of NPO’s that requires either the
-<big>ξ₂</big> or <big>β</big> outputs. Item f serves the purpose of handling higher power
-levels into and out of what effectively is a single (larger) NPO.
-<span class="pagenum" id="Page_63">[Pg 63]</span></p>
-
-<div class="figcenter">
- <img id="FIG_10E" src="images/i_070a.jpg" alt="" width="600" height="318" />
- <img src="images/i_070b.jpg" alt="" width="600" height="257" />
- <p class="f120 space-below2">Figure 10—Some possible networks of NPO’s</p>
-</div>
-
-<h3>CONCLUSION</h3>
-
-<p>The definition of self-organizing behavior suitably represented has
-permitted the use of Information Theoretic techniques to synthesize
-a (mathematical) mechanism for a self-organizing machine. Physical
-mechanization in the form of an NPO has been accomplished and has
-introduced the experimental phase of the program. From among the many
-items deserving of further study we may mention: more economical
-physical mechanization through introduction of modern technology;
-identification of networks of NPO’s with their group theoretic
-descriptions; analysis of the dimensionality of tasks which a SOM might
-be called on to simulate, and prototype SOM applications to related
-tasks. It is hoped that progress along these lines can be reported in
-the future.</p>
-<hr class="chap x-ebookmaker-drop" />
-
-<div class="chapter">
-<p><span class="pagenum" id="Page_64">[Pg 64]</span></p>
-<p class="f120 space-above1"><b>REFERENCES</b></p>
-</div>
-<table border="0" cellspacing="0" summary="REFERENCES" cellpadding="2" >
- <tbody><tr>
- <td id="REF_E_1" class="tdr">1.</td>
- <td class="tdl_ws1">Ścibor-Marchocki, Romuald I.,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">“A Topological Foundation for Self-Organization,”</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">Anaheim, California:Northrop Nortronics, NSS Report 2828,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">November 14, 1963</td>
- </tr><tr>
- <td id="REF_E_2" class="tdr_top">2.</td>
- <td class="tdl_ws1"><p class="no-indent">It is true that our definition is very similar to that proposed by
- Hawkins (reference 5). Compare for example his definition of learning
- machines (page 31 of reference 5). But the subsequent developments
- reviewed therein are different from the one we have followed.</p></td>
- </tr><tr>
- <td id="REF_E_3" class="tdr">3.</td>
- <td class="tdl_ws1">Ashby, W. R.,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">“The Set Theory of Mechanism and Homeostasis,”</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">Technical Report 7, University of Illinois,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">September 1962</td>
- </tr><tr>
- <td id="REF_E_4" class="tdr">4.</td>
- <td class="tdl_ws1">Ashby, W. R.,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">“Systems and Information,”</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1"><i>Transactions PTGME</i> <b>MIL-7</b>:94-97</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">(April-July, 1963)</td>
- </tr><tr>
- <td id="REF_E_5" class="tdr">5.</td>
- <td class="tdl_ws1">Hawkins, J. K.,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">“Self-Organizing Systems—A Review and Commentary,”</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1"><i>Proc. IRE</i>. <b>49</b>:31-48 (January 1961)</td>
- </tr><tr>
- <td id="REF_E_6" class="tdr">6.</td>
- <td class="tdl_ws1">Mesarovic, M. D.,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">“On Self Organizational Systems,”</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">Spartan Books, pp. 9-36, 1962</td>
- </tr><tr>
- <td id="REF_E_7" class="tdr">7.</td>
- <td class="tdl_ws1">Braverman, D.,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">“Learning Filters for Optimum Pattern Recognition,”</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1"><i>PGIT</i> <b>IT-8</b>:280-285 (July 1962)</td>
- </tr><tr>
- <td id="REF_E_8" class="tdr_top">8.</td>
- <td class="tdl_ws1"><p class="no-indent">We make the latter statement despite the fact that we employ a
- statistical treatment of self-organization. We may predict the
- performance of, for example, the NPO by using a statistical description,
- but it does not necessarily follow that the NPO computes statistics.</p></td>
- </tr><tr>
- <td id="REF_E_9" class="tdr">9.</td>
- <td class="tdl_ws1">McCulloch, W. S., and Pitts, W.,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">“A Logical Calculus of the Ideas Imminent in Nervous Activity,”</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1"><i>Bull-Math. Biophys</i> <b>5</b>:115 (1943)</td>
- </tr><tr>
- <td id="REF_E_10" class="tdr">10.</td>
- <td class="tdl_ws1">Newell, A., Shaw, J. C., and Simon, H. A.,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">“Empirical Explorations of the Logic Theory Machine:</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws2">A Case Study in Heuristic,”</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1"><i>Proc. WJCC</i>, pp. 218-230, 1957</td>
- </tr><tr>
- <td id="REF_E_10A" class="tdr_top">10a.</td>
- <td class="tdl_ws1"><p class="no-indent">The spaces W, X, Y, and Z are stochastic spaces; that is,
- each space is defined as the ordered pair (X,p(X)) where
- <big>p(X) = {p(x) ∋ x ∈ X}, p(x) ≥ 0, x ∈ X and ∫x p(x)dx = 1</big>.
- Such spaces possess a metrizable topology.</p></td>
- </tr><tr>
- <td id="REF_E_11" class="tdr_top">11.</td>
- <td class="tdl_ws1"><p class="no-indent">We use the following convention for probability distributions:
- if the arguments of p( ) are different, they are different
- functions, thus: <big>p(x) ≠ p(y)</big> even if <big>y = x</big>.</p></td>
- </tr><tr>
- <td id="REF_E_12" class="tdr_top">12.</td>
- <td class="tdl_ws1"><p class="no-indent">One can prove the existence of a metric directly but in order to
- perform the metrization the space has to be decomposed first. But
- decomposing a space without having a metric calls for a neat trick,
- accomplished (as far as we know) only by the method used by the SOM.</p></td>
-
- </tr><tr>
- <td id="REF_E_12A" class="tdr_top">12a.</td>
- <td class="tdl_ws1"><p class="no-indent">In this example we use a hemisphere;
- in general, it would be a spherical cap.</p></td>
-
- </tr>
- </tbody>
-</table>
-
-<hr class="chap x-ebookmaker-drop" />
-
-<div class="chapter">
-<p><span class="pagenum" id="Page_65">[Pg 65]</span></p>
-<h2 class="nobreak">A Topological Foundation for<br /> Self-Organization</h2>
-</div>
-
-<p class="f120"><b><span class="smcap">R. I. Ścibor-Marchocki</span></b></p>
-
-<p class="center space-below1"><i>Northrop Nortronics<br />
-Systems Support Department<br />Anaheim, California</i></p>
-
-<div class="blockquot">
-<p>It is shown that by the use of Information Theory, any metrizable
-topology may be metrized as an orthogonal Euclidean space (with a
-random Gaussian probability distribution) times a denumerable random
-cartesian product of irreducible (wrt direct product) denumerable
-groups. The necessary algorithm to accomplish this metrization from
-a statistical basis is presented. If such a basis is unavailable,
-a certain nilpotent projection operator has to be used instead, as
-is shown in detail in the companion paper. This operator possesses
-self-organizing features.</p>
-</div>
-
-<h3>INTRODUCTION</h3>
-
-<p>In the companion article<a id="FNanchor_8" href="#Footnote_8" class="fnanchor">[8]</a>
-we will define a self-organizing system as one which, after observing the input and output
-of an unknown phenomenon (transfer relation), organizes itself into a simulation of
-the unknown phenomenon.</p>
-
-<p>Within the mathematical model, the aforementioned phenomenon may be
-represented as a topological space thus omitting for the moment the
-(arbitrary) designation of input and output which, as will be shown,
-bears on the question of uniqueness. Hence, for the purpose of this
-paper, which emphasizes the mathematical foundation, an intelligent
-device is taken as one which carries out the task of studying a space
-and describing it.</p>
-
-<p>In keeping with the policy that one should not ask someone (or
-something) else to do a task that he could not do himself (at least in
-principle), let us consider how we would approach such a problem.</p>
-
-<p>In the first place, we have to select the space in which the problem is
-to be set. The most general space that we feel capable of tackling is
-a metrizable topology. On the other hand, anything less general would
-be unnecessarily restrictive. Thus, we choose a metrizable topological space.
-<span class="pagenum" id="Page_66">[Pg 66]</span></p>
-
-<p>As soon as we have made this choice, we regret it. In order to
-improve the situation somewhat, we show that there is no (additional)
-loss of generality in using an orthogonal Euclidean space times<a id="FNanchor_9" href="#Footnote_9" class="fnanchor">[9]</a>
-a denumerable random cartesian product of irreducible (wrt direct
-product) denumerable groups.</p>
-
-<p>This paper provides a survey of the problem and a method for solving
-it which is conceptually clear but not very practical. The companion
-paper<a id="FNanchor_10" href="#Footnote_10" class="fnanchor">[10]</a>
-provides a practical method for solving this problem by means
-of the successive use of a certain nilpotent projection operator.</p>
-
-<h3>METRIZATION</h3>
-
-<p>We start with a metrizable topological space. There are many
-equivalent axiomatizations of a metrizable topology; <i>e.g.</i>, see
-Kelley. Perhaps the easiest way to visualize a metrizable topology is
-to consider that one was given a metric space but that he lost his
-notes in which the exact form of the metric was written down. Thus one
-knows that he can do everything that he could in a metric space, if
-only he can figure out how.</p>
-
-<p>The “figuring out how” is by no means trivial. Here, it will be assumed
-that a cumulative probability distribution has been obtained on the
-space by one of the standard methods; bird in cage,<a id="FNanchor_11" href="#Footnote_11" class="fnanchor">[11]</a>
-Munroe I,<a id="FNanchor_12" href="#Footnote_12" class="fnanchor">[12]</a>
-Munroe II,<a id="FNanchor_13" href="#Footnote_13" class="fnanchor">[13]</a>
-ordering (see Halmos<a id="FNanchor_14" href="#Footnote_14" class="fnanchor">[14]</a>
-or Kelley<a id="FNanchor_15" href="#Footnote_15" class="fnanchor">[15]</a>).
-This cumulative probability distribution is a function on <big>X</big> onto the interval
-<big>[0,1]</big> of real numbers. The inverse of this function, which exists by the Radon
-Nikodym theorem, provides a mapping from the real interval onto the
-non-trivial portion of <big>X</big>. This mapping induces all of the pleasant properties
-of the real numbers on the space <big>X</big>: topological, metric, and ordering.</p>
-
-<p>Actually, it turns out that, especially if the dimensionality of
-the space is greater than one, the foregoing procedure not only
-provides one metrization, but many. Indeed, this lack of uniqueness
-is what makes the procedure exceedingly difficult. Only by imposing
-some additional conditions that result in the existence of a unique
-solution, does the problem become tractable.</p>
-
-<p>We choose to impose the additional condition that the resulting metric
-space be a Euclidean geometry with a rectangular coordinate system.
-<span class="pagenum" id="Page_67">[Pg 67]</span></p>
-
-<p>Even this always does not yield uniqueness, but we will show the
-additional restriction that will guarantee uniqueness after the
-necessary language is developed. Since all metrizations of a given
-metrizable topology are isomorphic, in the quotient class the
-orthogonal Euclidean geometry serves the purpose of being a convenient
-representative of the unique element resulting from a given metrizable
-topology.</p>
-
-<p>Furthermore, the same comment applies to the use of a Gaussian
-distribution as the probability distribution on this orthogonal
-Euclidean geometry. Namely, the random Gaussian distribution on an
-orthogonal Euclidean geometry is a convenient representative member of
-the equivalence class which maps into one element (stochastic space) of
-the quotient class.</p>
-
-<h3>Information Theory</h3>
-
-<p>Now, we will show that Information Theory provides the language
-necessary to describe the metrization procedure in detail.</p>
-
-<p>It is possible to introduce Information Theory axiomatically by
-a suitable generalization of the axioms<a id="FNanchor_16" href="#Footnote_16" class="fnanchor">[16]</a>
-in Feinstein.<a id="FNanchor_17" href="#Footnote_17" class="fnanchor">[17]</a>
-But to simplify the discussion here, we will use the less elegant but
-equivalent method of defining certain definite integrals. The
-probability density distribution p is defined from the cumulative
-probability distribution <big>P</big> by</p>
-
-<p class="center fontsize_130">P(X′) = ∫X′<sub><small>measurable ⊂ X</small></sub> p(x)dx.<span class="ws2">(1)</span></p>
-
-<p class="no-indent">Then the information rate H is defined as</p>
-
-<p class="center fontsize_130">H(X) = &nbsp;<big>-∫ₓ</big>p(x) &nbsp;ln&nbsp;κ&nbsp; p(x)dx<span class="ws2">(2)</span></p>
-
-<p class="no-indent">where kappa has (carries) the units of X. Finally,
-the channel rate R is defined as</p>
-
-<table class="fontsize_150" border="0" cellspacing="0" summary=" " cellpadding="0" >
- <tbody><tr>
- <td class="tdl">R</td>
- <td class="tdc">(⨀Xᵢ) =</td>
- <td class="tdc"><span class="fontsize_200">Σ</span></td>
- <td class="tdl">H(Xᵢ) - H(X),<span class="ws2"><small>(3)</small></span></td>
- </tr><tr>
- <td class="tdl">&nbsp;</td>
- <td class="tdl">&nbsp;&nbsp;<small>I</small></td>
- <td class="tdl">&nbsp;&nbsp;<small>I</small></td>
- <td class="tdl">&nbsp;</td>
- </tr>
- </tbody>
-</table>
-<p class="no-indent">where X is the denumerable<a id="FNanchor_18" href="#Footnote_18" class="fnanchor">[18]</a>
-cartesian product space</p>
-
-<table class="fontsize_130" border="0" cellspacing="0" summary=" " cellpadding="0" >
- <tbody><tr>
- <td class="tdl">X =&nbsp;</td>
- <td class="tdc">⨂</td>
- <td class="tdc">Xᵢ.<span class="ws2">(4)</span></td>
- </tr><tr>
- <td class="tdl">&nbsp;</td>
- <td class="tdl">&nbsp;<small>I</small></td>
- <td class="tdl">&nbsp;</td>
- </tr>
- </tbody>
-</table>
-<p><span class="pagenum" id="Page_68">[Pg 68]</span></p>
-
-<p class="no-indent">Next, we define the angle <big>Θ</big></p>
-
-<table class="fontsize_130" border="0" cellspacing="0" summary=" " cellpadding="0" >
- <tbody><tr>
- <td class="tdl">&nbsp;</td>
- <td class="tdl">&nbsp;</td>
- <td class="tdl">&nbsp;</td>
- <td class="tdl"><span class="fontsize_70">-R(⨀Xᵢ)</span></td>
- </tr><tr>
- <td class="tdl">&nbsp;</td>
- <td class="tdl">&nbsp;</td>
- <td class="tdl">&nbsp;</td>
- <td class="tdl"><span class="fontsize_70">&emsp;&nbsp; I</span></td>
- </tr><tr>
- <td class="tdl"><big>|</big>&nbsp;Θ</td>
- <td class="tdc"><big>(</big>⨀Xᵢ<big>)&nbsp;|</big> =&nbsp;</td>
- <td class="tdc">sin⁻¹<i>e</i></td>
- <td class="tdc"><span class="ws4">(5)</span></td>
- </tr><tr>
- <td class="tdl">&nbsp;</td>
- <td class="tdl">&nbsp; &nbsp;<small>I</small></td>
- <td class="tdl">&nbsp;</td>
- <td class="tdl">&nbsp;</td>
- </tr>
- </tbody>
-</table>
-<p class="no-indent">and the norm</p>
-
-<p class="f120">|X| = κ(2π<i>e</i>)⁻¹ᐟ² <i>e</i><sup>H(X)</sup>.<span class="ws4">(6)</span></p>
-
-<p class="no-indent">Now, if<a id="FNanchor_19" href="#Footnote_19" class="fnanchor">[19]</a>
-a statistically independent basis; <i>i.e.</i>, one for which</p>
-
-<table class="fontsize_130" border="0" cellspacing="0" summary=" " cellpadding="0" >
- <tbody><tr>
- <td class="tdl">&nbsp;</td>
- <td class="tdl">&nbsp; κ &nbsp;</td>
- <td class="tdl">&nbsp;</td>
- </tr><tr>
- <td class="tdl">R<big>(</big>⨀Xᵢ<big>)</big>&nbsp;</td>
- <td class="tdl">&nbsp; ≡ &nbsp;</td>
- <td class="tdl">constant,<span class="ws3">(7)</span></td>
- </tr><tr>
- <td class="tdl"><span class="ws2 fontsize_70">I</span></td>
- <td class="tdl">&nbsp;</td>
- <td class="tdl">&nbsp;</td>
- </tr>
- </tbody>
-</table>
-
-<p class="no-indent">can be provided in terms of one-dimensional
-components; <i>i.e.</i>, none of them can be decomposed further, then
-it is just the usual problem of diagonalization of a symmetric matrix
-by means of a congruence transformation to provide an orthogonal
-coordinate system. Furthermore, for uniqueness, we arrange the
-spectrum in decreasing order. Then, by means of the Radon Nikodym
-theorem applied to each of these one-dimensional axes, the probability
-distribution may be made; <i>e.g.</i>, Gaussian, if desired. Thus, we
-obtain the promised orthogonal Euclidean space.</p>
-
-<h3>Channel</h3>
-
-<p>At this time we can state the remaining additional condition required
-that a decomposition be unique. The index space I has to be partitioned
-into exactly two parts, say <big>I′</big> and <big>I″</big>; <i>i.e.</i>,</p>
-
-<p class="f120">I′ ∪ I″ = I<span class="ws3">(8)</span></p>
-
-<p class="f120">I′ ∩ I″ = φ,<span class="ws3">&nbsp;&nbsp;&nbsp;</span></p>
-
-<p class="no-indent">such that</p>
-
-<p class="f120"><b>dim</b>(X′) &nbsp;= &nbsp;<b>dim</b>(X″),<span class="ws2">(9)</span></p>
-
-<p class="no-indent">where</p>
-
-<table class="fontsize_130" border="0" cellspacing="0" summary=" " cellpadding="0" >
- <tbody><tr>
- <td class="tdl">X′ =&nbsp;</td>
- <td class="tdl">⨂Xᵢ<span class="ws3">(10)</span></td>
- </tr><tr>
- <td class="tdl">&nbsp;</td>
- <td class="tdl fontsize_70">&nbsp;I′</td>
- </tr><tr>
- <td class="tdl">X″ =&nbsp;</td>
- <td class="tdl">⨂Xᵢ.</td>
- </tr><tr>
- <td class="tdl">&nbsp;</td>
- <td class="tdl fontsize_70">&nbsp;I″</td>
- </tr>
- </tbody>
-</table>
-
-<p><span class="pagenum" id="Page_69">[Pg 69]</span>
-(If <big><b>dim</b> (X)</big> is odd, then we have to cheat a little by putting in an
-extra random dummy dimension.) And then the decomposition of the space</p>
-
-<table class="fontsize_130" border="0" cellspacing="0" summary=" " cellpadding="0" >
- <tbody><tr>
- <td class="tdl">X =&nbsp;</td>
- <td class="tdl">⨂Xᵢ<span class="ws3">(11)</span></td>
- </tr><tr>
- <td class="tdl">&nbsp;</td>
- <td class="tdl fontsize_70">&nbsp;I</td>
- </tr>
- </tbody>
-</table>
-
-<p class="no-indent">has to be carried out so that this partitioning is preserved.
-Since this partitioning is arbitrary (as far as the mathematics is
-concerned), it is obvious that a space which is not partitioned will
-have many (equivalent) decompositions. On the other hand, if the
-partitioning is into more than two parts, then the existence of a
-decomposition is not guaranteed.</p>
-
-<p>A slight penalty has to be paid for the use of this partitioning,
-namely: instead of eventually obtaining a random cartesian product of
-one-dimensional spaces, we obtain an extended channel (with random
-input) of single-dimensional channels. It is obvious that if we were
-to drop the partitioning temporarily, each such single-dimensional
-channel would be further decomposed into two random components. This
-decomposition is not unique. But one of these equivalent decompositions
-is particularly convenient; namely, that decomposition where we take
-the component out of the original X′ and that which is random to it,
-say V. This V (as well as the cartesian product of all such V’s,
-which of necessity are random) is called the linearly additive noise.
-The name “linearly additive” is justified because it is just the
-statistical concept isomorphic to the linear addition of vectors in
-orthogonal Euclidean geometry. (The proof of this last statement is not
-completed as yet.)</p>
-
-<h3>Denumerable Space</h3>
-
-<p>The procedure for this decomposition was worded to de-emphasize the
-possible presence of a denumerable (component of the) space. Such a
-component may be given outright; otherwise, it results if the space was
-not simply connected. Any denumerable space is zero dimensional, as may
-be verified easily from the full information theoretic definition of
-dimensionality.</p>
-
-<p>The obvious way of disposing of a denumerable space is to use the
-conventional mapping that converts a Stieltjes to a Lebesque
-integral, using fixed length segments. (It can be shown that H is
-invariant under such a mapping.) Unfortunately, while this mapping
-followed by a repetition of the preceding procedure will always solve a
-<span class="pagenum" id="Page_70">[Pg 70]</span>
-given problem (no new<a id="FNanchor_20" href="#Footnote_20" class="fnanchor">[20]</a>
-denumerable component <i>need</i> be generated on the second pass),
-little insight is provided into the structure of the resulting space.
-On the other hand, because channels under cascading constitute a group,
-any such denumerable space is a representation of a denumerable group.</p>
-
-<h3>SUMMARY</h3>
-
-<p>In summary, the original metrizable topological space was decomposed
-into an orthogonal Euclidean space times<a id="FNanchor_21" href="#Footnote_21" class="fnanchor">[21]</a>
-a denumerable random cartesian product of irreducible (wrt direct
-product) denumerable groups. Thus, since any individual component of a
-random cartesian product may be studied independently of the others,
-all that one needs to study is: (1) a Gaussian distribution on a single
-real axis and (2) the irreducible denumerable groups.</p>
-
-<p>Finally, it should be emphasized that there are only these two ways
-of decomposing a metrizable topology; (1) if a (statistical) basis
-is given, use the diagonalization of a symmetric matrix algorithm
-described earlier (and given in detail in the three channels in cascade
-problem), and (2) otherwise use a suitable network of the NPO’s with
-n₀=1. Of course, any hybrid of these two methods may be employed as well.</p>
-
-<hr class="chap x-ebookmaker-drop" />
-
-<div class="chapter">
-<p><span class="pagenum" id="Page_71">[Pg 71]</span></p>
-<h2 class="nobreak">On Functional Neuron Modeling</h2>
-</div>
-
-<p class="f120"><b><span class="smcap">C. E. Hendrix</span></b></p>
-
-<p class="center space-below1"><i>Space-General Corporation<br />
-El Monte, California</i></p>
-
-<p>There are two very compelling reasons why mathematical and physical
-models of the neuron should be built. Model building, while widely
-used in the physical sciences, has been largely neglected in biology.
-However, there can be little doubt that building neuron models
-will increase our understanding of the function of real neurons,
-if experience in the physical sciences is any guide. Secondly,
-neuron models are extremely interesting in their own right as new
-technological devices. Hence, the interest in, and the reason for
-symposia on self-organizing systems.</p>
-
-<p>We should turn our attention to the properties of real neurons, and
-see which of them are the most important ones for us to imitate.
-Obviously, we cannot hope to imitate <i>all</i> the properties of a
-living neuron, since that would require a complete simulation of a
-living, metabolizing cell, and a highly specialized one at that; but
-we can select those functional properties which we feel are the most
-important, and then try to simulate those.</p>
-
-<p>The most dramatic aspect of neuron function is, of course, the axon
-discharge. It is this which gives the neuron its “all-or-nothing”
-character, and it is this which provides it with a means for
-propagating its output pulses over a distance. <a href="#REF_G_1">Hodgkin and Huxley (1)</a>
-have developed a very complete description of this action. Their model
-is certainly without peer in describing the nature of the real neuron.</p>
-
-<p>On the technological side, Cranes’ “neuristors” <a href="#REF_G_2">(2)</a> represent a class
-of devices which imitate the axonal discharge in a gross sort of way,
-without all the subtle nuances of the Hodgkin-Huxley model. Crane has
-shown that neuristors can be combined to yield the various Boolean
-functions needed in a computer.</p>
-
-<p>However, interesting as such models of the axon are, there is some
-question as to their importance in the development of self-organizing
-systems. The pulse generation, “all-or-nothing” part of the axon
-behavior could just as well be simulated by a “one-shot” trigger
-circuit. The transmission characteristic of the axon is, after all,
-only Nature’s way of sending a signal from here to there. It is an
-<span class="pagenum" id="Page_72">[Pg 72]</span>
-admirable solution to the problem, when one considers that it
-evolved, and still works, in a bath of salt water. There seems little
-point, however, in a hardware designer limiting himself in this way,
-especially if he has an adequate supply of insulated copper wire.</p>
-
-<p>If the transmission characteristic of the axon is deleted, the
-properties of the neuron which seem to be the most important in the
-synthesis of self-organizing systems are:</p>
-
-<div class="blockquot">
-<p>a. The neuron responds to a stimulus with an electrical pulse of
-standard size and shape. If the stimulus continues, the pulses occur
-at regular intervals with the rate of occurrence dependent on the
-intensity of stimulation.</p>
-
-<p>b. There is a threshold of stimulation. If the intensity of the
-stimulus is below this threshold, the neuron does not fire.</p>
-
-<p>c. The neuron is capable of temporal and spatial integration. Many
-subthreshold stimuli arriving at the neuron from different sources, or
-at slightly different times, can add up to a sufficient level to fire
-the neuron.</p>
-
-<p>d. Some inputs are excitatory, some are inhibitory.</p>
-
-<p>e. There is a refractory period. Once fired, there is a subsequent
-period during which the neuron cannot be fired again, no matter how
-large the stimulus. This places an upper limit on the pulse rate of any
-particular neuron.</p>
-
-<p>f. The neuron can learn. This property is conjectural in living
-neurons, since it appears that at the present time learning has not
-been clearly demonstrated in isolated living neurons. However, the
-learning property is basic to all self-organizing models.</p>
-</div>
-
-<p>Neuron models with the above characteristics have been built, although
-none seem to have incorporated <i>all</i> of them in a single model.
-<a href="#REF_G_3">Harman (3)</a> at Bell Labs has built neuron models which
-have the characteristics (a) through (e), with which he has built extremely
-interesting devices which simulate portions of the peripheral neuron
-system.</p>
-
-<p>Various attempts at learning elements have been made, perhaps best
-exemplified by those of <a href="#REF_G_4">Widrow (4)</a>. These devices are
-capable of “learning,” but are static, and lack all the temporal characteristics
-listed in (a) through (e). Such devices can be used to deal with
-temporal patterns only by a mapping technique, in which a temporal
-pattern is converted to a spatial one.</p>
-
-<p>Having listed which seem to be the important properties of a neuron,
-it is possible to synthesize a simple model which has all of them.
-<span class="pagenum" id="Page_73">[Pg 73]</span></p>
-
-<p>A number of input stimuli are fed to the neuron through a resistive
-summing network which establishes the threshold and accomplishes
-spatial integration. The voltage at the summing junction triggers a
-“one-shot” circuit, which, by its very nature, accomplishes pulse
-generation and exhibits temporal integration and a refractory period.
-The polarity of an individual input determines whether it shall be
-excitatory or inhibitory. This much of the circuitry is very similar to
-Harmon’s model.</p>
-
-<p>Learning is postulated to take place in the following way: when the
-neuron fires, an outside influence (the environment, or a “trainer”)
-determines whether or not the result of firing was desirable or not.
-If it was desirable, the threshold of the neuron is lowered, making
-it easier to fire the next time. If the result was not desirable, the
-threshold is raised, making it more difficult for the neuron to fire
-the next time.</p>
-
-<p>In a self-organizing system, many model neurons would be
-interconnected. A “punish-reward” (P-R) signal would be connected to
-all neurons in common. However, means would be provided for only those
-which have recently fired to be susceptible to the effects of the P-R
-signal. Therefore, only those which had taken part in a recent response
-are modified. This idea is due to <a href="#REF_G_5">Stewart (5)</a>, who applies
-it to his electrochemical devices instead of to an electronic device.</p>
-
-<p>The mechanization of the circuitry is rather straight-forward. A
-portion of the output of the pulse generator is routed through a
-“pulse-stretcher” or short-term memory which temporarily records the
-fact that the neuron has recently fired. The pulse-stretcher output
-controls a gate, which either accepts or rejects the P-R signal. The
-P-R signal can take on only three values, a positive level, zero, or
-a negative level, depending on whether the signal is “punish,” “no
-action,” or “reward.” Finally, the gate output controls a variable
-resistor, which is part of the resistive summing network. <a href="#FIG_1G">Figure 1</a>
-is a block diagram of the complete model.</p>
-
-<p>Note that this device differs from the usual “Perceptron” configuration
-in that the threshold resistor is the only variable element, instead
-of having each input resistor a variable weighting element. This
-simplification could lead to a situation where, to prepare a specified
-task, more single-variable neurons would be required than would
-multivariable ones. This possible disadvantage is partially, at least,
-offset by the very simple control algorithm which is contained in the
-design of the model, and is not the matter of great concern which it
-seems to be for most multivariable models.
-<span class="pagenum" id="Page_74">[Pg 74]</span></p>
-
-<div class="figcenter">
- <img id="FIG_1G" src="images/i_081.jpg" alt="" width="600" height="350" />
- <p class="f120 space-below2">Figure 1—Block diagram of neuron model</p>
-</div>
-
-<p>Hand simulations of the action of this type of model suggest that a
-certain amount of randomness would be desirable. It appears that a
-self-organizing system built of these elements, and of sufficient
-complexity to be interesting, would have a fair number of recirculating
-loops, so that spontaneous activity would be maintained in the absence
-of input stimulus. If this is the case, then randomness could easily
-be introduced by adding a small amount of noise from a random noise
-generator to the signal on the P-R bus. Thus, any neurons which
-spontaneously fire would be continually having their thresholds modified.</p>
-
-<p>The mechanization of the model is not particularly complex, and can
-be estimated as follows: The one-shot pulse generator would require
-two transistors, the pulse stretcher one more. The bi-directional gate
-would require a transistor and at least two diodes.</p>
-
-<p>Several candidates for the electrically-controllable variable
-resistor are available <a href="#REF_G_6">(6)</a>. Particularly good candidates appear to
-be the “Memistor” or plating cell developed by <a href="#REF_G_7">Widrow (7)</a>, the
-solid state version of it by <a href="#REF_G_8">Vendelin (8)</a>, and the “solion”
-<a href="#REF_G_9">(9)</a>. All are electrochemical devices in which
-the resistance between two terminals is controlled by the net charge
-flow through a third terminal. All are adaptable to this particular circuit.</p>
-
-<p>Of the three, however, the solion appears at first glance to have the
-most promise in that its resistance is of the order of a few thousand
-ohms (rather than the few ohms of the plating cells) which is more
-compatible with ordinary solid-state circuitry. Solions have the
-disadvantage that they can stand only very low voltages (less than 1
-volt) and in their present form require extra bias potentials. If these
-difficulties can be overcome, they offer considerable promise.
-<span class="pagenum" id="Page_75">[Pg 75]</span></p>
-
-<p>In summary, it appears that a rather simple neuron model can be built
-which can mimic most of the important functions of real neurons. A
-system built of these could be punished or rewarded by an observer,
-so that it could be trained to give specified responses to specified
-stimuli. In some cases, the observer could be simply the environment,
-so that the system would learn directly from experience, and would be
-therefore a self-organizing system.</p>
-
-<p class="f120 space-above1"><b>REFERENCES</b></p>
-<table border="0" cellspacing="0" summary="REFERENCES" cellpadding="2" >
- <tbody><tr>
- <td id="REF_G_1" class="tdr">1.</td>
- <td class="tdl_ws1">Hodgkin, A. L., and Huxley, A. L.,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">“A Quantitative Description of Membrane Current and its
- Application to Conduction and Excitation in Nerve,”</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1"><i>J. Physiol.</i> <b>117</b>:500-544 (August 1952)</td>
- </tr><tr>
- <td id="REF_G_2" class="tdr">2.</td>
- <td class="tdl_ws1">Crane, H. D.,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">“Neuristor—A Novel Device and System Concept,”</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1"><i>Proc. IRE</i> <b>50</b>:2048-2060 (Oct. 1962)</td>
- </tr><tr>
- <td id="REF_G_3" class="tdr">3.</td>
- <td class="tdl_ws1">Harmon, L. D., Levinson, J., and Van Bergeijk, W. A.,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">“Analog Models of Neural Mechanism,”</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1"><i>IRE Trans. on Information Theory</i> <b>IT-8</b>:107-112</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">(Feb. 1962)</td>
- </tr><tr>
- <td id="REF_G_4" class="tdr">4.</td>
- <td class="tdl_ws1">Widrow, B., and Hoff, M. E.,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">“Adaptive Switching Circuits,”</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">Stanford Electronics Lab Tech Report 1553-1, June 1960</td>
- </tr><tr>
- <td id="REF_G_5" class="tdr">5.</td>
- <td class="tdl_ws1">Stewart, R. M.,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">“Electrochemical Wave Interactions and Extensive
- Field Effects in Excitable Cellular Structures,”</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">First Pasadena Invitational Symposium on Self-Organizing Systems,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">Calif. Institute of Technology, Pasadena, Calif., 14 Nov. 1963</td>
- </tr><tr>
- <td id="REF_G_6" class="tdr">6.</td>
- <td class="tdl_ws1">Nagy, G.,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">“A Survey of Analog Memory Devices,”</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1"><i>IEEE Trans. on Electronic Cmptrs.</i> EC-12:388-393 (Aug. 1963)</td>
- </tr><tr>
- <td id="REF_G_7" class="tdr">7.</td>
- <td class="tdl_ws1">Widrow, B.,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">“An Adaptive Adaline Neuron Using Chemical Memistors,”</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">Stanford Electronics Lab Tech Report 1553-2, Oct. 1960</td>
- </tr><tr>
- <td id="REF_G_8" class="tdr">8.</td>
- <td class="tdl_ws1">Vendelin, G. D.,</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">“A Solid State Adaptive Component,”</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">Stanford Electronics Lab Tech Report 1853-1, Jan. 1963</td>
- </tr><tr>
- <td id="REF_G_9" class="tdr">9.</td>
- <td class="tdl_ws1">“Solion Principles of Electrochemistry and Low-Power
- Electrochemical Devices,”</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">Dept. of Comm., Office of Tech. Serv. <b>PB</b> 131931</td>
- </tr><tr>
- <td class="tdr">&nbsp;</td>
- <td class="tdl_ws1">(U. S. Naval Ord. Lab., Silver Spring, Md., Aug. 1958)</td>
- </tr>
- </tbody>
-</table>
-
-<hr class="chap x-ebookmaker-drop" />
-
-<div class="chapter">
-<p><span class="pagenum" id="Page_76">[Pg 76]</span></p>
-
-<h2 class="nobreak">Selection of Parameters for<br />
-Neural Net Simulations<a id="FNanchor_22" href="#Footnote_22" class="fnanchor"><small>[22]</small></a></h2>
-</div>
-
-<p class="f120"><b><span class="smcap">R. K. Overton</span></b></p>
-
-<p class="center space-below1"><i>Autonetics Research Center<br />
-Anaheim, California</i></p>
-
-<p>Research of high quality has been presented at this Symposium. Of
-particular interest to me were the reports of the Aeronutronic group
-and the Librascope group. The Aeronutronic group was commendably
-systematic in its investigations of different arrangements of linear
-threshold elements, and the Librascope data, presenting the effects of
-attaching different values to the parameters of simulated neurons, are
-both systematic and interesting.</p>
-
-<p>Unfortunately, however, interest in such research can obscure a more
-fundamental question which seems to merit study. That question concerns
-the parameters, or attributes, which describe the simulated neuron.
-Specifically, which parameters or attributes should be selected for
-simulation? (For example, should a period of supernormal sensitivity be
-simulated following an absolutely refractory period?)</p>
-
-<p>Some selection obviously has to be made. Librascope, which is
-trying to simulate neurons more or less faithfully, plans to build
-a net of ten simulated neurons. In contrast, General Dynamics/Fort
-Worth, with roughly the same degree of effort, is working with 3900
-unfaithfully-simulated neurons. This comparison is not a criticism
-of either group; the Librascope team has simply selected many more
-parameters for simulation than has the General Dynamics group. Each
-can make the selections it prefers, because the parameters of real
-neurons which are necessary and sufficient for learning have not been
-exhaustively identified.</p>
-
-<p>From the point of view of one whose interests include real neurons,
-this lack of identification is unfortunate. I once wrote a book which
-included some guesses about the essential attributes of neurons. Since
-that time, many neuron simulation programs have been written. But these
-programs, although interesting and worthwhile in their own right, have
-done little to answer the question of the necessary parameters. That
-is, they do not make much better guesses possible. And yet better
-guesses would also make for more “intelligent” machines.</p>
-
-<hr class="chap x-ebookmaker-drop" />
-
-<div class="chapter">
-<p><span class="pagenum" id="Page_77">[Pg 77]</span></p>
-<h2 class="nobreak">INDEX OF INVITED PARTICIPANTS</h2>
-</div>
-
-<table border="0" cellspacing="0" summary="PARTICIPANTS" cellpadding="2" >
- <tbody><tr>
- <td class="tdl">MICHAEL ARBIB</td>
- <td class="tdl_ws1">Massachusetts Institute of Technology</td>
- </tr><tr>
- <td class="tdl">ROBERT H. ASENDORF</td>
- <td class="tdl_ws1">Hughes Research Laboratories/ Malibu</td>
- </tr><tr>
- <td class="tdl">J. A. DALY</td>
- <td class="tdl_ws1">Astropower/Newport Beach</td>
- </tr><tr>
- <td class="tdl">GEORGE DeFLORIO</td>
- <td class="tdl_ws1">System Development Corp./Santa Monica</td>
- </tr><tr>
- <td class="tdl">DEREK H. FENDER</td>
- <td class="tdl_ws1">California Institute of Technology</td>
- </tr><tr>
- <td class="tdl">LEONARD FRIEDMAN</td>
- <td class="tdl_ws1">Space Technology Labs./Redondo Beach</td>
- </tr><tr>
- <td class="tdl">JAMES EMMETT GARVEY</td>
- <td class="tdl_ws1">ONR/Pasadena</td>
- </tr><tr>
- <td class="tdl">THOMAS L. GRETTENBERG</td>
- <td class="tdl_ws1">California Institute of Technology</td>
- </tr><tr>
- <td class="tdl">HAROLD HAMILTON</td>
- <td class="tdl_ws1">Librascope/Glendale</td>
- </tr><tr>
- <td class="tdl">JOSEPH HAWKINS</td>
- <td class="tdl_ws1">Aeronutronic/Newport Beach</td>
- </tr><tr>
- <td class="tdl">CHARLES HENDRIX</td>
- <td class="tdl_ws1">Space-General Corp./El Monte</td>
- </tr><tr>
- <td class="tdl">R. D. JOSEPH</td>
- <td class="tdl_ws1">Astropower/Newport Beach</td>
- </tr><tr>
- <td class="tdl">PETER A. KLEYN</td>
- <td class="tdl_ws1">Nortronics/Anaheim</td>
- </tr><tr>
- <td class="tdl">JOHN KUHN</td>
- <td class="tdl_ws1">Space-General Corp./El Monte</td>
- </tr><tr>
- <td class="tdl">FRANK LEHAN</td>
- <td class="tdl_ws1">Space-General Corp./El Monte</td>
- </tr><tr>
- <td class="tdl">EDWIN LEWIS</td>
- <td class="tdl_ws1">Librascope/Glendale</td>
- </tr><tr>
- <td class="tdl">PETER C. LOCKEMANN</td>
- <td class="tdl_ws1">California Institute of Technology</td>
- </tr><tr>
- <td class="tdl">GILBERT D. McCANN</td>
- <td class="tdl_ws1">California Institute of Technology
- <span class="pagenum" id="Page_78">[Pg 78]</span></td>
- </tr><tr>
- <td class="tdl">C. J. MUNCIE</td>
- <td class="tdl_ws1">Aeronutronic/Newport Beach</td>
- </tr><tr>
- <td class="tdl">C. OVERMIER</td>
- <td class="tdl_ws1">Nortronics/Anaheim</td>
- </tr><tr>
- <td class="tdl">RICHARD K. OVERTON</td>
- <td class="tdl_ws1">Autonetics/Anaheim</td>
- </tr><tr>
- <td class="tdl">DIANE RAMSEY</td>
- <td class="tdl_ws1">Astropower/Newport Beach</td>
- </tr><tr>
- <td class="tdl">RICHARD REISS</td>
- <td class="tdl_ws1">Librascope/Glendale</td>
- </tr><tr>
- <td class="tdl">R. I. ŚCIBOR-MARCHOCKI</td>
- <td class="tdl_ws1">Nortronics/Anaheim</td>
- </tr><tr>
- <td class="tdl">JAMES J. SPILKER</td>
- <td class="tdl_ws1">Philco/Palo Alto</td>
- </tr><tr>
- <td class="tdl">ROBERT M. STEWART</td>
- <td class="tdl_ws1">Space-General Corp./El Monte</td>
- </tr><tr>
- <td class="tdl">HENNIG STIEVE</td>
- <td class="tdl_ws1">California Institute of Technology</td>
- </tr><tr>
- <td class="tdl">RICHARD TEW</td>
- <td class="tdl_ws1">Space-General Corp./El Monte</td>
- </tr><tr>
- <td class="tdl">JOHN THORSEN</td>
- <td class="tdl_ws1">University of California/Los Angeles</td>
- </tr><tr>
- <td class="tdl">RICHARD VINETZ</td>
- <td class="tdl_ws1">Librascope/Glendale</td>
- </tr><tr>
- <td class="tdl">CHRISTOPH von CAMPENHAUSEN</td>
- <td class="tdl_ws1">California Institute of Technology</td>
- </tr><tr>
- <td class="tdl">DAVID VOWLES</td>
- <td class="tdl_ws1">California Institute of Technology</td>
- </tr><tr>
- <td class="tdl">HORST WOLF</td>
- <td class="tdl_ws1">Astropower/Newport Beach</td>
- </tr>
- </tbody>
-</table>
-
-<hr class="chap x-ebookmaker-drop" />
-<p class="center space-above2">U.S. GOVERNMENT PRINTING OFFICE: 1966 O—205-502</p>
-<hr class="chap x-ebookmaker-drop" />
-
-<div class="footnotes">
-<p class="f150"><b>Footnotes:</b></p>
-
-<div class="footnote"><p>
-<a id="Footnote_1" href="#FNanchor_1" class="label">[1]</a>
-For review articles see: <a href="#REF_B_13">Lillie (13)</a>,
-<a href="#REF_B_6">Franck (6)</a>.</p></div>
-
-<div class="footnote"><p>
-<a id="Footnote_2" href="#FNanchor_2" class="label">[2]</a>
-The operation of this machine is described in substantially greater
-detail in J. J. Spilker, Jr., D. D. Luby, R. D. Lawhorn, “Adaptive
-Binary Waveform Detection,” Philco Western Development Laboratories,
-Communication Sciences Department, TR #75, December 1963.</p></div>
-
-<div class="footnote"><p class="no-indent">
-<a id="Footnote_3" href="#FNanchor_3" class="label">[3]</a>
-F. M. Glaser, “Signal Detection by Adaptive Filters,”
-<i>IRE Trans. Information Theory</i>, pp. 87-90; April 1961.</p></div>
-
-<div class="footnote"><p class="no-indent">
-<a id="Footnote_4" href="#FNanchor_4" class="label">[4]</a>
-P. W. Cooper, “The Hypersphere in Pattern Recognition,”
-<i>Information and Control</i>, pp. 324-346; December 1962.</p></div>
-
-<div class="footnote"><p class="no-indent">
-<a id="Footnote_5" href="#FNanchor_5" class="label">[5]</a>
-Observed from Oscillogram</p></div>
-
-<div class="footnote"><p class="no-indent">
-<a id="Footnote_6" href="#FNanchor_6" class="label">[6]</a>
-Computed</p></div>
-
-<div class="footnote"><p class="no-indent">
-<a id="Footnote_7" href="#FNanchor_7" class="label">[7]</a>
-Observed from Oscillogram</p></div>
-
-<div class="footnote"><p class="no-indent">
-<a id="Footnote_8" href="#FNanchor_8" class="label">[8]</a>
-Kleyn, P. A., “Conceptual Design of Self-Organizing
-Machines,” Anaheim, California:Northrop Nortronics, NSS Report 2832,
-Nov. 14, 1963.</p></div>
-
-<div class="footnote"><p class="no-indent">
-<a id="Footnote_9" href="#FNanchor_9" class="label">[9]</a>
-Random cartesian product.</p></div>
-
-<div class="footnote"><p class="no-indent">
-<a id="Footnote_10" href="#FNanchor_10" class="label">[10]</a>
-Kleyn, P. A., “Conceptual Design of Self-Organizing
-Machines,” Anaheim, California:Northrop Nortronics, NSS Report 2832,
-Nov. 14, 1963.</p></div>
-
-<div class="footnote"><p class="no-indent">
-<a id="Footnote_11" href="#FNanchor_11" class="label">[11]</a>
-Harman, W. W., “Principles of the Statistical Theory of
-Communication,” New York, New York:McGraw-Hill, 1963.</p></div>
-
-<div class="footnote"><p class="no-indent">
-<a id="Footnote_12" href="#FNanchor_12" class="label">[12]</a>
-Munroe, M. E., “Introduction to Measure and Integration,”
-Cambridge, Mass.:Addison-Wesley, 1953.</p></div>
-
-<div class="footnote"><p class="no-indent">
-<a id="Footnote_13" href="#FNanchor_13" class="label">[13]</a>
-Munroe, M. E., “Introduction to Measure and Integration,”
-Cambridge, Mass.:Addison-Wesley, 1953.</p></div>
-
-<div class="footnote"><p class="no-indent">
-<a id="Footnote_14" href="#FNanchor_14" class="label">[14]</a>
-Halmos, P. R., “Measure Theory,” Princeton, New Jersey:D.
-Van Nostrand Co., Inc., 1950.</p></div>
-
-<div class="footnote"><p class="no-indent">
-<a id="Footnote_15" href="#FNanchor_15" class="label">[15]</a>
-Kelley, J. L., “General Topology,” Princeton, New
-Jersey:D. Van Nostrand Co., Inc., 1955.</p></div>
-
-<div class="footnote"><p class="no-indent">
-<a id="Footnote_16" href="#FNanchor_16" class="label">[16]</a>
-Feinstein uses his axioms only in finite space X;
-<i>i.e.</i>, card(X) &lt; K₀.</p></div>
-
-<div class="footnote"><p class="no-indent">
-<a id="Footnote_17" href="#FNanchor_17" class="label">[17]</a>
-Feinstein, A., “Foundations of Information Theory,”
-New York, New York: McGraw-Hill, 1958.</p></div>
-
-<div class="footnote"><p class="no-indent">
-<a id="Footnote_18" href="#FNanchor_18" class="label">[18]</a>
-If I is infinite, certain precautions have to be exercised.</p></div>
-
-<div class="footnote"><p class="no-indent">
-<a id="Footnote_19" href="#FNanchor_19" class="label">[19]</a>
-This “if” is the catch that makes all methods of metrization of a space
-of dimensionality higher than one impractical, except the method of
-successive projections upon unit spheres centered at the center of
-gravity. The method of using that nilpotent projection operator is
-described in the companion paper(see footnote <a href="#Page_65">page 65</a>).</p></div>
-
-<div class="footnote"><p class="no-indent">
-<a id="Footnote_20" href="#FNanchor_20" class="label">[20]</a>
-Only non-cyclic irreducible (wrt direct product) denumerable
-group components of the old denumerable space will remain.</p></div>
-
-<div class="footnote"><p class="no-indent">
-<a id="Footnote_21" href="#FNanchor_21" class="label">[21]</a>
-Random cartesian product.</p></div>
-
-<div class="footnote"><p class="no-indent">
-<a id="Footnote_22" href="#FNanchor_22" class="label">[22]</a>
-This paper, submitted after the Symposium, represents a more detailed
-presentation of some of the issues raised in the discussion sessions
-at the Symposium and hence, constitutes a worthwhile addition to the
-Proceedings.</p></div>
-</div>
-
-<div class="transnote bbox space-above2">
-<p class="f120 space-above1">Transcriber’s Notes:</p>
-<hr class="r5" />
-<p class="indent">The illustrations have been moved so that they do not break up
- paragraphs and so that they are next to the text they illustrate.</p>
-<p class="indent">Typographical and punctuation errors have been silently corrected.</p>
-<p class="indent">A heavy bar on top of a letter indicates a vector,
- e.g. <span class="bt2">M</span> means “the vector M”.</p>
-</div>
-
-<div style='display:block; margin-top:4em'>*** END OF THE PROJECT GUTENBERG EBOOK SELF-ORGANIZING SYSTEMS, 1963 ***</div>
-<div style='text-align:left'>
-
-<div style='display:block; margin:1em 0'>
-Updated editions will replace the previous one&#8212;the old editions will
-be renamed.
-</div>
-
-<div style='display:block; margin:1em 0'>
-Creating the works from print editions not protected by U.S. copyright
-law means that no one owns a United States copyright in these works,
-so the Foundation (and you!) can copy and distribute it in the United
-States without permission and without paying copyright
-royalties. Special rules, set forth in the General Terms of Use part
-of this license, apply to copying and distributing Project
-Gutenberg&#8482; electronic works to protect the PROJECT GUTENBERG&#8482;
-concept and trademark. Project Gutenberg is a registered trademark,
-and may not be used if you charge for an eBook, except by following
-the terms of the trademark license, including paying royalties for use
-of the Project Gutenberg trademark. If you do not charge anything for
-copies of this eBook, complying with the trademark license is very
-easy. You may use this eBook for nearly any purpose such as creation
-of derivative works, reports, performances and research. Project
-Gutenberg eBooks may be modified and printed and given away--you may
-do practically ANYTHING in the United States with eBooks not protected
-by U.S. copyright law. Redistribution is subject to the trademark
-license, especially commercial redistribution.
-</div>
-
-<div style='margin:0.83em 0; font-size:1.1em; text-align:center'>START: FULL LICENSE<br />
-<span style='font-size:smaller'>THE FULL PROJECT GUTENBERG LICENSE<br />
-PLEASE READ THIS BEFORE YOU DISTRIBUTE OR USE THIS WORK</span>
-</div>
-
-<div style='display:block; margin:1em 0'>
-To protect the Project Gutenberg&#8482; mission of promoting the free
-distribution of electronic works, by using or distributing this work
-(or any other work associated in any way with the phrase &#8220;Project
-Gutenberg&#8221;), you agree to comply with all the terms of the Full
-Project Gutenberg&#8482; License available with this file or online at
-www.gutenberg.org/license.
-</div>
-
-<div style='display:block; font-size:1.1em; margin:1em 0; font-weight:bold'>
-Section 1. General Terms of Use and Redistributing Project Gutenberg&#8482; electronic works
-</div>
-
-<div style='display:block; margin:1em 0'>
-1.A. By reading or using any part of this Project Gutenberg&#8482;
-electronic work, you indicate that you have read, understand, agree to
-and accept all the terms of this license and intellectual property
-(trademark/copyright) agreement. If you do not agree to abide by all
-the terms of this agreement, you must cease using and return or
-destroy all copies of Project Gutenberg&#8482; electronic works in your
-possession. If you paid a fee for obtaining a copy of or access to a
-Project Gutenberg&#8482; electronic work and you do not agree to be bound
-by the terms of this agreement, you may obtain a refund from the person
-or entity to whom you paid the fee as set forth in paragraph 1.E.8.
-</div>
-
-<div style='display:block; margin:1em 0'>
-1.B. &#8220;Project Gutenberg&#8221; is a registered trademark. It may only be
-used on or associated in any way with an electronic work by people who
-agree to be bound by the terms of this agreement. There are a few
-things that you can do with most Project Gutenberg&#8482; electronic works
-even without complying with the full terms of this agreement. See
-paragraph 1.C below. There are a lot of things you can do with Project
-Gutenberg&#8482; electronic works if you follow the terms of this
-agreement and help preserve free future access to Project Gutenberg&#8482;
-electronic works. See paragraph 1.E below.
-</div>
-
-<div style='display:block; margin:1em 0'>
-1.C. The Project Gutenberg Literary Archive Foundation (&#8220;the
-Foundation&#8221; or PGLAF), owns a compilation copyright in the collection
-of Project Gutenberg&#8482; electronic works. Nearly all the individual
-works in the collection are in the public domain in the United
-States. If an individual work is unprotected by copyright law in the
-United States and you are located in the United States, we do not
-claim a right to prevent you from copying, distributing, performing,
-displaying or creating derivative works based on the work as long as
-all references to Project Gutenberg are removed. Of course, we hope
-that you will support the Project Gutenberg&#8482; mission of promoting
-free access to electronic works by freely sharing Project Gutenberg&#8482;
-works in compliance with the terms of this agreement for keeping the
-Project Gutenberg&#8482; name associated with the work. You can easily
-comply with the terms of this agreement by keeping this work in the
-same format with its attached full Project Gutenberg&#8482; License when
-you share it without charge with others.
-</div>
-
-<div style='display:block; margin:1em 0'>
-1.D. The copyright laws of the place where you are located also govern
-what you can do with this work. Copyright laws in most countries are
-in a constant state of change. If you are outside the United States,
-check the laws of your country in addition to the terms of this
-agreement before downloading, copying, displaying, performing,
-distributing or creating derivative works based on this work or any
-other Project Gutenberg&#8482; work. The Foundation makes no
-representations concerning the copyright status of any work in any
-country other than the United States.
-</div>
-
-<div style='display:block; margin:1em 0'>
-1.E. Unless you have removed all references to Project Gutenberg:
-</div>
-
-<div style='display:block; margin:1em 0'>
-1.E.1. The following sentence, with active links to, or other
-immediate access to, the full Project Gutenberg&#8482; License must appear
-prominently whenever any copy of a Project Gutenberg&#8482; work (any work
-on which the phrase &#8220;Project Gutenberg&#8221; appears, or with which the
-phrase &#8220;Project Gutenberg&#8221; is associated) is accessed, displayed,
-performed, viewed, copied or distributed:
-</div>
-
-<blockquote>
- <div style='display:block; margin:1em 0'>
- This eBook is for the use of anyone anywhere in the United States and most
- other parts of the world at no cost and with almost no restrictions
- whatsoever. You may copy it, give it away or re-use it under the terms
- of the Project Gutenberg License included with this eBook or online
- at <a href="https://www.gutenberg.org">www.gutenberg.org</a>. If you
- are not located in the United States, you will have to check the laws
- of the country where you are located before using this eBook.
- </div>
-</blockquote>
-
-<div style='display:block; margin:1em 0'>
-1.E.2. If an individual Project Gutenberg&#8482; electronic work is
-derived from texts not protected by U.S. copyright law (does not
-contain a notice indicating that it is posted with permission of the
-copyright holder), the work can be copied and distributed to anyone in
-the United States without paying any fees or charges. If you are
-redistributing or providing access to a work with the phrase &#8220;Project
-Gutenberg&#8221; associated with or appearing on the work, you must comply
-either with the requirements of paragraphs 1.E.1 through 1.E.7 or
-obtain permission for the use of the work and the Project Gutenberg&#8482;
-trademark as set forth in paragraphs 1.E.8 or 1.E.9.
-</div>
-
-<div style='display:block; margin:1em 0'>
-1.E.3. If an individual Project Gutenberg&#8482; electronic work is posted
-with the permission of the copyright holder, your use and distribution
-must comply with both paragraphs 1.E.1 through 1.E.7 and any
-additional terms imposed by the copyright holder. Additional terms
-will be linked to the Project Gutenberg&#8482; License for all works
-posted with the permission of the copyright holder found at the
-beginning of this work.
-</div>
-
-<div style='display:block; margin:1em 0'>
-1.E.4. Do not unlink or detach or remove the full Project Gutenberg&#8482;
-License terms from this work, or any files containing a part of this
-work or any other work associated with Project Gutenberg&#8482;.
-</div>
-
-<div style='display:block; margin:1em 0'>
-1.E.5. Do not copy, display, perform, distribute or redistribute this
-electronic work, or any part of this electronic work, without
-prominently displaying the sentence set forth in paragraph 1.E.1 with
-active links or immediate access to the full terms of the Project
-Gutenberg&#8482; License.
-</div>
-
-<div style='display:block; margin:1em 0'>
-1.E.6. You may convert to and distribute this work in any binary,
-compressed, marked up, nonproprietary or proprietary form, including
-any word processing or hypertext form. However, if you provide access
-to or distribute copies of a Project Gutenberg&#8482; work in a format
-other than &#8220;Plain Vanilla ASCII&#8221; or other format used in the official
-version posted on the official Project Gutenberg&#8482; website
-(www.gutenberg.org), you must, at no additional cost, fee or expense
-to the user, provide a copy, a means of exporting a copy, or a means
-of obtaining a copy upon request, of the work in its original &#8220;Plain
-Vanilla ASCII&#8221; or other form. Any alternate format must include the
-full Project Gutenberg&#8482; License as specified in paragraph 1.E.1.
-</div>
-
-<div style='display:block; margin:1em 0'>
-1.E.7. Do not charge a fee for access to, viewing, displaying,
-performing, copying or distributing any Project Gutenberg&#8482; works
-unless you comply with paragraph 1.E.8 or 1.E.9.
-</div>
-
-<div style='display:block; margin:1em 0'>
-1.E.8. You may charge a reasonable fee for copies of or providing
-access to or distributing Project Gutenberg&#8482; electronic works
-provided that:
-</div>
-
-<div style='margin-left:0.7em;'>
- <div style='text-indent:-0.7em'>
- &bull; You pay a royalty fee of 20% of the gross profits you derive from
- the use of Project Gutenberg&#8482; works calculated using the method
- you already use to calculate your applicable taxes. The fee is owed
- to the owner of the Project Gutenberg&#8482; trademark, but he has
- agreed to donate royalties under this paragraph to the Project
- Gutenberg Literary Archive Foundation. Royalty payments must be paid
- within 60 days following each date on which you prepare (or are
- legally required to prepare) your periodic tax returns. Royalty
- payments should be clearly marked as such and sent to the Project
- Gutenberg Literary Archive Foundation at the address specified in
- Section 4, &#8220;Information about donations to the Project Gutenberg
- Literary Archive Foundation.&#8221;
- </div>
-
- <div style='text-indent:-0.7em'>
- &bull; You provide a full refund of any money paid by a user who notifies
- you in writing (or by e-mail) within 30 days of receipt that s/he
- does not agree to the terms of the full Project Gutenberg&#8482;
- License. You must require such a user to return or destroy all
- copies of the works possessed in a physical medium and discontinue
- all use of and all access to other copies of Project Gutenberg&#8482;
- works.
- </div>
-
- <div style='text-indent:-0.7em'>
- &bull; You provide, in accordance with paragraph 1.F.3, a full refund of
- any money paid for a work or a replacement copy, if a defect in the
- electronic work is discovered and reported to you within 90 days of
- receipt of the work.
- </div>
-
- <div style='text-indent:-0.7em'>
- &bull; You comply with all other terms of this agreement for free
- distribution of Project Gutenberg&#8482; works.
- </div>
-</div>
-
-<div style='display:block; margin:1em 0'>
-1.E.9. If you wish to charge a fee or distribute a Project
-Gutenberg&#8482; electronic work or group of works on different terms than
-are set forth in this agreement, you must obtain permission in writing
-from the Project Gutenberg Literary Archive Foundation, the manager of
-the Project Gutenberg&#8482; trademark. Contact the Foundation as set
-forth in Section 3 below.
-</div>
-
-<div style='display:block; margin:1em 0'>
-1.F.
-</div>
-
-<div style='display:block; margin:1em 0'>
-1.F.1. Project Gutenberg volunteers and employees expend considerable
-effort to identify, do copyright research on, transcribe and proofread
-works not protected by U.S. copyright law in creating the Project
-Gutenberg&#8482; collection. Despite these efforts, Project Gutenberg&#8482;
-electronic works, and the medium on which they may be stored, may
-contain &#8220;Defects,&#8221; such as, but not limited to, incomplete, inaccurate
-or corrupt data, transcription errors, a copyright or other
-intellectual property infringement, a defective or damaged disk or
-other medium, a computer virus, or computer codes that damage or
-cannot be read by your equipment.
-</div>
-
-<div style='display:block; margin:1em 0'>
-1.F.2. LIMITED WARRANTY, DISCLAIMER OF DAMAGES - Except for the &#8220;Right
-of Replacement or Refund&#8221; described in paragraph 1.F.3, the Project
-Gutenberg Literary Archive Foundation, the owner of the Project
-Gutenberg&#8482; trademark, and any other party distributing a Project
-Gutenberg&#8482; electronic work under this agreement, disclaim all
-liability to you for damages, costs and expenses, including legal
-fees. YOU AGREE THAT YOU HAVE NO REMEDIES FOR NEGLIGENCE, STRICT
-LIABILITY, BREACH OF WARRANTY OR BREACH OF CONTRACT EXCEPT THOSE
-PROVIDED IN PARAGRAPH 1.F.3. YOU AGREE THAT THE FOUNDATION, THE
-TRADEMARK OWNER, AND ANY DISTRIBUTOR UNDER THIS AGREEMENT WILL NOT BE
-LIABLE TO YOU FOR ACTUAL, DIRECT, INDIRECT, CONSEQUENTIAL, PUNITIVE OR
-INCIDENTAL DAMAGES EVEN IF YOU GIVE NOTICE OF THE POSSIBILITY OF SUCH
-DAMAGE.
-</div>
-
-<div style='display:block; margin:1em 0'>
-1.F.3. LIMITED RIGHT OF REPLACEMENT OR REFUND - If you discover a
-defect in this electronic work within 90 days of receiving it, you can
-receive a refund of the money (if any) you paid for it by sending a
-written explanation to the person you received the work from. If you
-received the work on a physical medium, you must return the medium
-with your written explanation. The person or entity that provided you
-with the defective work may elect to provide a replacement copy in
-lieu of a refund. If you received the work electronically, the person
-or entity providing it to you may choose to give you a second
-opportunity to receive the work electronically in lieu of a refund. If
-the second copy is also defective, you may demand a refund in writing
-without further opportunities to fix the problem.
-</div>
-
-<div style='display:block; margin:1em 0'>
-1.F.4. Except for the limited right of replacement or refund set forth
-in paragraph 1.F.3, this work is provided to you &#8216;AS-IS&#8217;, WITH NO
-OTHER WARRANTIES OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT
-LIMITED TO WARRANTIES OF MERCHANTABILITY OR FITNESS FOR ANY PURPOSE.
-</div>
-
-<div style='display:block; margin:1em 0'>
-1.F.5. Some states do not allow disclaimers of certain implied
-warranties or the exclusion or limitation of certain types of
-damages. If any disclaimer or limitation set forth in this agreement
-violates the law of the state applicable to this agreement, the
-agreement shall be interpreted to make the maximum disclaimer or
-limitation permitted by the applicable state law. The invalidity or
-unenforceability of any provision of this agreement shall not void the
-remaining provisions.
-</div>
-
-<div style='display:block; margin:1em 0'>
-1.F.6. INDEMNITY - You agree to indemnify and hold the Foundation, the
-trademark owner, any agent or employee of the Foundation, anyone
-providing copies of Project Gutenberg&#8482; electronic works in
-accordance with this agreement, and any volunteers associated with the
-production, promotion and distribution of Project Gutenberg&#8482;
-electronic works, harmless from all liability, costs and expenses,
-including legal fees, that arise directly or indirectly from any of
-the following which you do or cause to occur: (a) distribution of this
-or any Project Gutenberg&#8482; work, (b) alteration, modification, or
-additions or deletions to any Project Gutenberg&#8482; work, and (c) any
-Defect you cause.
-</div>
-
-<div style='display:block; font-size:1.1em; margin:1em 0; font-weight:bold'>
-Section 2. Information about the Mission of Project Gutenberg&#8482;
-</div>
-
-<div style='display:block; margin:1em 0'>
-Project Gutenberg&#8482; is synonymous with the free distribution of
-electronic works in formats readable by the widest variety of
-computers including obsolete, old, middle-aged and new computers. It
-exists because of the efforts of hundreds of volunteers and donations
-from people in all walks of life.
-</div>
-
-<div style='display:block; margin:1em 0'>
-Volunteers and financial support to provide volunteers with the
-assistance they need are critical to reaching Project Gutenberg&#8482;&#8217;s
-goals and ensuring that the Project Gutenberg&#8482; collection will
-remain freely available for generations to come. In 2001, the Project
-Gutenberg Literary Archive Foundation was created to provide a secure
-and permanent future for Project Gutenberg&#8482; and future
-generations. To learn more about the Project Gutenberg Literary
-Archive Foundation and how your efforts and donations can help, see
-Sections 3 and 4 and the Foundation information page at www.gutenberg.org.
-</div>
-
-<div style='display:block; font-size:1.1em; margin:1em 0; font-weight:bold'>
-Section 3. Information about the Project Gutenberg Literary Archive Foundation
-</div>
-
-<div style='display:block; margin:1em 0'>
-The Project Gutenberg Literary Archive Foundation is a non-profit
-501(c)(3) educational corporation organized under the laws of the
-state of Mississippi and granted tax exempt status by the Internal
-Revenue Service. The Foundation&#8217;s EIN or federal tax identification
-number is 64-6221541. Contributions to the Project Gutenberg Literary
-Archive Foundation are tax deductible to the full extent permitted by
-U.S. federal laws and your state&#8217;s laws.
-</div>
-
-<div style='display:block; margin:1em 0'>
-The Foundation&#8217;s business office is located at 809 North 1500 West,
-Salt Lake City, UT 84116, (801) 596-1887. Email contact links and up
-to date contact information can be found at the Foundation&#8217;s website
-and official page at www.gutenberg.org/contact
-</div>
-
-<div style='display:block; font-size:1.1em; margin:1em 0; font-weight:bold'>
-Section 4. Information about Donations to the Project Gutenberg Literary Archive Foundation
-</div>
-
-<div style='display:block; margin:1em 0'>
-Project Gutenberg&#8482; depends upon and cannot survive without widespread
-public support and donations to carry out its mission of
-increasing the number of public domain and licensed works that can be
-freely distributed in machine-readable form accessible by the widest
-array of equipment including outdated equipment. Many small donations
-($1 to $5,000) are particularly important to maintaining tax exempt
-status with the IRS.
-</div>
-
-<div style='display:block; margin:1em 0'>
-The Foundation is committed to complying with the laws regulating
-charities and charitable donations in all 50 states of the United
-States. Compliance requirements are not uniform and it takes a
-considerable effort, much paperwork and many fees to meet and keep up
-with these requirements. We do not solicit donations in locations
-where we have not received written confirmation of compliance. To SEND
-DONATIONS or determine the status of compliance for any particular state
-visit <a href="https://www.gutenberg.org/donate/">www.gutenberg.org/donate</a>.
-</div>
-
-<div style='display:block; margin:1em 0'>
-While we cannot and do not solicit contributions from states where we
-have not met the solicitation requirements, we know of no prohibition
-against accepting unsolicited donations from donors in such states who
-approach us with offers to donate.
-</div>
-
-<div style='display:block; margin:1em 0'>
-International donations are gratefully accepted, but we cannot make
-any statements concerning tax treatment of donations received from
-outside the United States. U.S. laws alone swamp our small staff.
-</div>
-
-<div style='display:block; margin:1em 0'>
-Please check the Project Gutenberg web pages for current donation
-methods and addresses. Donations are accepted in a number of other
-ways including checks, online payments and credit card donations. To
-donate, please visit: www.gutenberg.org/donate
-</div>
-
-<div style='display:block; font-size:1.1em; margin:1em 0; font-weight:bold'>
-Section 5. General Information About Project Gutenberg&#8482; electronic works
-</div>
-
-<div style='display:block; margin:1em 0'>
-Professor Michael S. Hart was the originator of the Project
-Gutenberg&#8482; concept of a library of electronic works that could be
-freely shared with anyone. For forty years, he produced and
-distributed Project Gutenberg&#8482; eBooks with only a loose network of
-volunteer support.
-</div>
-
-<div style='display:block; margin:1em 0'>
-Project Gutenberg&#8482; eBooks are often created from several printed
-editions, all of which are confirmed as not protected by copyright in
-the U.S. unless a copyright notice is included. Thus, we do not
-necessarily keep eBooks in compliance with any particular paper
-edition.
-</div>
-
-<div style='display:block; margin:1em 0'>
-Most people start at our website which has the main PG search
-facility: <a href="https://www.gutenberg.org">www.gutenberg.org</a>.
-</div>
-
-<div style='display:block; margin:1em 0'>
-This website includes information about Project Gutenberg&#8482;,
-including how to make donations to the Project Gutenberg Literary
-Archive Foundation, how to help produce our new eBooks, and how to
-subscribe to our email newsletter to hear about new eBooks.
-</div>
-
-</div>
-
-</body>
-</html>
diff --git a/old/66286-h/images/cbl-3.jpg b/old/66286-h/images/cbl-3.jpg
deleted file mode 100644
index 6adebc6..0000000
--- a/old/66286-h/images/cbl-3.jpg
+++ /dev/null
Binary files differ
diff --git a/old/66286-h/images/cover.jpg b/old/66286-h/images/cover.jpg
deleted file mode 100644
index d38ee56..0000000
--- a/old/66286-h/images/cover.jpg
+++ /dev/null
Binary files differ
diff --git a/old/66286-h/images/i_010a.jpg b/old/66286-h/images/i_010a.jpg
deleted file mode 100644
index 5cd470e..0000000
--- a/old/66286-h/images/i_010a.jpg
+++ /dev/null
Binary files differ
diff --git a/old/66286-h/images/i_010b.jpg b/old/66286-h/images/i_010b.jpg
deleted file mode 100644
index 5bf516a..0000000
--- a/old/66286-h/images/i_010b.jpg
+++ /dev/null
Binary files differ
diff --git a/old/66286-h/images/i_010c.jpg b/old/66286-h/images/i_010c.jpg
deleted file mode 100644
index 7349a75..0000000
--- a/old/66286-h/images/i_010c.jpg
+++ /dev/null
Binary files differ
diff --git a/old/66286-h/images/i_010d.jpg b/old/66286-h/images/i_010d.jpg
deleted file mode 100644
index 1b00fc1..0000000
--- a/old/66286-h/images/i_010d.jpg
+++ /dev/null
Binary files differ
diff --git a/old/66286-h/images/i_011a.jpg b/old/66286-h/images/i_011a.jpg
deleted file mode 100644
index 5ea8334..0000000
--- a/old/66286-h/images/i_011a.jpg
+++ /dev/null
Binary files differ
diff --git a/old/66286-h/images/i_011b.jpg b/old/66286-h/images/i_011b.jpg
deleted file mode 100644
index 30eabbf..0000000
--- a/old/66286-h/images/i_011b.jpg
+++ /dev/null
Binary files differ
diff --git a/old/66286-h/images/i_011c.jpg b/old/66286-h/images/i_011c.jpg
deleted file mode 100644
index ee230f9..0000000
--- a/old/66286-h/images/i_011c.jpg
+++ /dev/null
Binary files differ
diff --git a/old/66286-h/images/i_011d.jpg b/old/66286-h/images/i_011d.jpg
deleted file mode 100644
index c3aeebe..0000000
--- a/old/66286-h/images/i_011d.jpg
+++ /dev/null
Binary files differ
diff --git a/old/66286-h/images/i_011e.jpg b/old/66286-h/images/i_011e.jpg
deleted file mode 100644
index 6318f39..0000000
--- a/old/66286-h/images/i_011e.jpg
+++ /dev/null
Binary files differ
diff --git a/old/66286-h/images/i_012a.jpg b/old/66286-h/images/i_012a.jpg
deleted file mode 100644
index 7f6157d..0000000
--- a/old/66286-h/images/i_012a.jpg
+++ /dev/null
Binary files differ
diff --git a/old/66286-h/images/i_012b.jpg b/old/66286-h/images/i_012b.jpg
deleted file mode 100644
index de090f0..0000000
--- a/old/66286-h/images/i_012b.jpg
+++ /dev/null
Binary files differ
diff --git a/old/66286-h/images/i_012c.jpg b/old/66286-h/images/i_012c.jpg
deleted file mode 100644
index 66f8316..0000000
--- a/old/66286-h/images/i_012c.jpg
+++ /dev/null
Binary files differ
diff --git a/old/66286-h/images/i_013a.jpg b/old/66286-h/images/i_013a.jpg
deleted file mode 100644
index 0e404bf..0000000
--- a/old/66286-h/images/i_013a.jpg
+++ /dev/null
Binary files differ
diff --git a/old/66286-h/images/i_013b.jpg b/old/66286-h/images/i_013b.jpg
deleted file mode 100644
index fb87e85..0000000
--- a/old/66286-h/images/i_013b.jpg
+++ /dev/null
Binary files differ
diff --git a/old/66286-h/images/i_016.jpg b/old/66286-h/images/i_016.jpg
deleted file mode 100644
index bf9cc36..0000000
--- a/old/66286-h/images/i_016.jpg
+++ /dev/null
Binary files differ
diff --git a/old/66286-h/images/i_017.jpg b/old/66286-h/images/i_017.jpg
deleted file mode 100644
index ff28e07..0000000
--- a/old/66286-h/images/i_017.jpg
+++ /dev/null
Binary files differ
diff --git a/old/66286-h/images/i_019.jpg b/old/66286-h/images/i_019.jpg
deleted file mode 100644
index fea9fd2..0000000
--- a/old/66286-h/images/i_019.jpg
+++ /dev/null
Binary files differ
diff --git a/old/66286-h/images/i_035.jpg b/old/66286-h/images/i_035.jpg
deleted file mode 100644
index 1351b7c..0000000
--- a/old/66286-h/images/i_035.jpg
+++ /dev/null
Binary files differ
diff --git a/old/66286-h/images/i_037a.jpg b/old/66286-h/images/i_037a.jpg
deleted file mode 100644
index 1df2044..0000000
--- a/old/66286-h/images/i_037a.jpg
+++ /dev/null
Binary files differ
diff --git a/old/66286-h/images/i_037b.jpg b/old/66286-h/images/i_037b.jpg
deleted file mode 100644
index c2540bb..0000000
--- a/old/66286-h/images/i_037b.jpg
+++ /dev/null
Binary files differ
diff --git a/old/66286-h/images/i_038.jpg b/old/66286-h/images/i_038.jpg
deleted file mode 100644
index f176f47..0000000
--- a/old/66286-h/images/i_038.jpg
+++ /dev/null
Binary files differ
diff --git a/old/66286-h/images/i_039a.jpg b/old/66286-h/images/i_039a.jpg
deleted file mode 100644
index a3531d4..0000000
--- a/old/66286-h/images/i_039a.jpg
+++ /dev/null
Binary files differ
diff --git a/old/66286-h/images/i_039b.jpg b/old/66286-h/images/i_039b.jpg
deleted file mode 100644
index d0ef117..0000000
--- a/old/66286-h/images/i_039b.jpg
+++ /dev/null
Binary files differ
diff --git a/old/66286-h/images/i_041a.jpg b/old/66286-h/images/i_041a.jpg
deleted file mode 100644
index fc1a6fa..0000000
--- a/old/66286-h/images/i_041a.jpg
+++ /dev/null
Binary files differ
diff --git a/old/66286-h/images/i_041b.jpg b/old/66286-h/images/i_041b.jpg
deleted file mode 100644
index 53ad1ad..0000000
--- a/old/66286-h/images/i_041b.jpg
+++ /dev/null
Binary files differ
diff --git a/old/66286-h/images/i_041c.jpg b/old/66286-h/images/i_041c.jpg
deleted file mode 100644
index d680c36..0000000
--- a/old/66286-h/images/i_041c.jpg
+++ /dev/null
Binary files differ
diff --git a/old/66286-h/images/i_055.jpg b/old/66286-h/images/i_055.jpg
deleted file mode 100644
index 2ea7c7f..0000000
--- a/old/66286-h/images/i_055.jpg
+++ /dev/null
Binary files differ
diff --git a/old/66286-h/images/i_056a.jpg b/old/66286-h/images/i_056a.jpg
deleted file mode 100644
index 5ce7d7d..0000000
--- a/old/66286-h/images/i_056a.jpg
+++ /dev/null
Binary files differ
diff --git a/old/66286-h/images/i_056b.jpg b/old/66286-h/images/i_056b.jpg
deleted file mode 100644
index 3cf7a82..0000000
--- a/old/66286-h/images/i_056b.jpg
+++ /dev/null
Binary files differ
diff --git a/old/66286-h/images/i_057a.jpg b/old/66286-h/images/i_057a.jpg
deleted file mode 100644
index ec46d14..0000000
--- a/old/66286-h/images/i_057a.jpg
+++ /dev/null
Binary files differ
diff --git a/old/66286-h/images/i_057b.jpg b/old/66286-h/images/i_057b.jpg
deleted file mode 100644
index cf5307c..0000000
--- a/old/66286-h/images/i_057b.jpg
+++ /dev/null
Binary files differ
diff --git a/old/66286-h/images/i_057c.jpg b/old/66286-h/images/i_057c.jpg
deleted file mode 100644
index 691fe9f..0000000
--- a/old/66286-h/images/i_057c.jpg
+++ /dev/null
Binary files differ
diff --git a/old/66286-h/images/i_057d.jpg b/old/66286-h/images/i_057d.jpg
deleted file mode 100644
index e555300..0000000
--- a/old/66286-h/images/i_057d.jpg
+++ /dev/null
Binary files differ
diff --git a/old/66286-h/images/i_057e.jpg b/old/66286-h/images/i_057e.jpg
deleted file mode 100644
index 065bdb8..0000000
--- a/old/66286-h/images/i_057e.jpg
+++ /dev/null
Binary files differ
diff --git a/old/66286-h/images/i_060.jpg b/old/66286-h/images/i_060.jpg
deleted file mode 100644
index 9114212..0000000
--- a/old/66286-h/images/i_060.jpg
+++ /dev/null
Binary files differ
diff --git a/old/66286-h/images/i_061a.jpg b/old/66286-h/images/i_061a.jpg
deleted file mode 100644
index e413120..0000000
--- a/old/66286-h/images/i_061a.jpg
+++ /dev/null
Binary files differ
diff --git a/old/66286-h/images/i_061b.jpg b/old/66286-h/images/i_061b.jpg
deleted file mode 100644
index cea11b1..0000000
--- a/old/66286-h/images/i_061b.jpg
+++ /dev/null
Binary files differ
diff --git a/old/66286-h/images/i_062.jpg b/old/66286-h/images/i_062.jpg
deleted file mode 100644
index 44556ae..0000000
--- a/old/66286-h/images/i_062.jpg
+++ /dev/null
Binary files differ
diff --git a/old/66286-h/images/i_064.jpg b/old/66286-h/images/i_064.jpg
deleted file mode 100644
index f97e989..0000000
--- a/old/66286-h/images/i_064.jpg
+++ /dev/null
Binary files differ
diff --git a/old/66286-h/images/i_065.jpg b/old/66286-h/images/i_065.jpg
deleted file mode 100644
index bb4e108..0000000
--- a/old/66286-h/images/i_065.jpg
+++ /dev/null
Binary files differ
diff --git a/old/66286-h/images/i_066.jpg b/old/66286-h/images/i_066.jpg
deleted file mode 100644
index 327d4b0..0000000
--- a/old/66286-h/images/i_066.jpg
+++ /dev/null
Binary files differ
diff --git a/old/66286-h/images/i_067a.jpg b/old/66286-h/images/i_067a.jpg
deleted file mode 100644
index 9aa5a71..0000000
--- a/old/66286-h/images/i_067a.jpg
+++ /dev/null
Binary files differ
diff --git a/old/66286-h/images/i_067b.jpg b/old/66286-h/images/i_067b.jpg
deleted file mode 100644
index 0795e67..0000000
--- a/old/66286-h/images/i_067b.jpg
+++ /dev/null
Binary files differ
diff --git a/old/66286-h/images/i_070a.jpg b/old/66286-h/images/i_070a.jpg
deleted file mode 100644
index e6b0539..0000000
--- a/old/66286-h/images/i_070a.jpg
+++ /dev/null
Binary files differ
diff --git a/old/66286-h/images/i_070b.jpg b/old/66286-h/images/i_070b.jpg
deleted file mode 100644
index 1bdf7f9..0000000
--- a/old/66286-h/images/i_070b.jpg
+++ /dev/null
Binary files differ
diff --git a/old/66286-h/images/i_081.jpg b/old/66286-h/images/i_081.jpg
deleted file mode 100644
index b31ea94..0000000
--- a/old/66286-h/images/i_081.jpg
+++ /dev/null
Binary files differ