Introduction To The Theory Of Neural Computation

by Hertz,John A.

• ISBN13: 9780201515602
• ISBN10: 0201515601
• Format: Nonspecific Binding
• Copyright: 1991-06-24
• Publisher: CRC Press

Summary

This book is a comprehensive introduction to the neural network models currently under intensive study for computational applications. It is a detailed, logically-developed treatment that covers the theory and uses of collective computational networks, including associative memory, feed forward networks, and unsupervised learning. It also provides coverage of neural network applications in a variety of problems of both theoretical and practical interest.

Table of Contents

Series Foreword	xiii
L. M. Simmons, Jr.
Foreword	xv
Jack Cowan
Foreword	xvii
Christof Koch
Preface	xix
Introduction	1	(10)
Inspiration from Neuroscience	2	(4)
History	6	(2)
The Issues	8	(3)
The Hopfield Model	11	(32)
The Associative Memory Problem	11	(2)
The Model	13	(12)
Statistical Mechanics of Magnetic Systems	25	(7)
Stochastic Networks	32	(3)
Capacity of the Stochastic Network	35	(8)
Extensions of the Hopfield Model	43	(28)
Variations on the Hopfield Model	43	(6)
Correlated Patterns	49	(4)
Continuous-Valued Units	53	(5)
Hardware Implementations	58	(5)
Temporal Sequences of Patterns	63	(8)
Optimization Problems	71	(18)
The Weighted Matching Problem	72	(4)
The Travelling Salesman Problem	76	(3)
Graph Bipartitioning	79	(2)
Optimization Problems in Image Processing	81	(8)
Simple Perceptrons	89	(26)
Feed-Forward Networks	90	(2)
Threshold Units	92	(8)
Proof of Convergence of the Perceptron Learning Rule	100	(2)
Linear Units	102	(5)
Nonlinear Units	107	(3)
Stochastic Units	110	(1)
Capacity of the Simple Perceptron	111	(4)
Multi-Layer Networks	115	(48)
Back-Propagation	115	(5)
Variations on Back-Propagation	120	(10)
Examples and Applications	130	(11)
Performance of Multi-Layer Feed-Forward Networks	141	(6)
A Theoretical Framework for Generalization	147	(9)
Optimal Network Architectures	156	(7)
Recurrent Networks	163	(34)
Boltzmann Machines	163	(9)
Recurrent Back-Propagation	172	(4)
Learning Time Sequences	176	(12)
Reinforcement Learning	188	(9)
Unsupervised Hebbian Learning	197	(20)
Unsupervised Learning	197	(2)
One Linear Unit	199	(5)
Principal Component Analysis	204	(6)
Self-Organizing Feature Extraction	210	(7)
Unsupervised Competitive Learning	217	(34)
Simple Competitive Learning	218	(5)
Examples and Applications of Competitive Learning	223	(5)
Adaptive Resonance Theory	228	(4)
Feature Mapping	232	(8)
Theory of Feature Mapping	240	(4)
The Travelling Salesman Problem	244	(2)
Hybrid Learning Schemes	246	(5)
Formal Statistical Mechanics of Neural Networks	251	(24)
The Hopfield Model	251	(14)
Gardner Theory of the Connections	265	(10)
APPENDIX Statistical Mechanics	275	(6)
A.1 The Boltzmann-Gibbs Distribution	275	(2)
A.2 Free Energy and Entropy	277	(2)
A.3 Stochastic Dynamics	279	(2)
Bibliography	281	(26)
Subject Index	307	(14)
Author Index	321