Elements of Artificial Neural Networks

Transparency masters, including abbreviated text and figures for the entire book, are available for instructors using the text.

Preface

xiii

Introduction

1

(42)

History of Neural Networks

4

(3)

Structure and Function of a Single Neuron

7

(9)

Biological neurons

7

(2)

Artificial neuron models

9

(7)

Neural Net Architectures

16

(6)

Fully connected networks

17

(1)

Layered networks

18

(1)

Acyclic networks

18

(2)

Feedforward networks

20

(1)

Modular neural networks

21

(1)

Neural Learning

22

(2)

Correlation learning

22

(1)

Competitive learning

22

(1)

Feedback-based weight adaptation

23

(1)

What Can Neural Networks Be Used for?

24

(11)

Classification

25

(1)

Clustering

25

(2)

Vector quantization

27

(1)

Pattern association

27

(2)

Function approximation

29

(2)

Forecasting

31

(1)

Control applications

32

(2)

Optimization

34

(1)

Search

34

(1)

Evaluation of Networks

35

(3)

Quality of results

36

(1)

Generalizability

37

(1)

Computational resources

38

(1)

Implementation

38

(1)

Conclusion

39

(2)

Exercises

41

(2)

Supervised Learning: Single-Layer Networks

43

(22)

Perceptrons

43

(2)

Linear Separability

45

(1)

Perceptron Training Algorithm

46

(6)

Termination criterion

50

(1)

Choice of learning rate

50

(1)

Non-numeric inputs

51

(1)

Guarantee of Success

52

(2)

Modifications

54

(7)

Pocket algorithm

55

(2)

Adalines

57

(3)

Multiclass discrimination

60

(1)

Conclusion

61

(1)

Exercises

62

(3)

Supervised Learning: Multilayer Networks I

65

(46)

Multilevel Discrimination

66

(1)

Preliminaries

67

(3)

Architecture

67

(1)

Objectives

68

(2)

Backpropagation Algorithm

70

(9)

Setting the Parameter Values

79

(9)

Initialization of weights

79

(1)

Frequency of weight updates

80

(1)

Choice of learning rate

81

(2)

Momentum

83

(1)

Generalizability

84

(1)

Number of hidden layers and nodes

85

(1)

Number of samples

86

(2)

Theoretical Results

88

(5)

Cover's theorem

88

(2)

Representations of functions

90

(1)

Approximations of functions

91

(2)

Accelerating the Learning Process

93

(5)

Quickprop algorithm

93

(1)

Conjugate gradient

94

(4)

Applications

98

(7)

Weaning from mechanically assisted ventilation

98

(2)

Classification of myoelectric signals

100

(1)

Forecasting commodity prices

101

(2)

Controlling a gantry crane

103

(2)

Conclusion

105

(1)

Exercises

106

(5)

Supervised Learning: Multilayer Networks II

111

(46)

Madalines

111

(5)

Adaptive Multilayer Networks

116

(20)

Network pruning algorithms

116

(2)

Marchand's algorithm

118

(5)

Upstart algorithm

123

(5)

Neural Tree

128

(2)

Cascade correlation

130

(3)

Tiling algorithm

133

(3)

Prediction Networks

136

(5)

Recurrent networks

136

(3)

Feedforward networks for forecasting

139

(2)

Radial Basis Functions

141

(8)

Polynomial Networks

149

(4)

Regularization

153

(1)

Conclusion

154

(1)

Exercises

155

(2)

Unsupervised Learning

157

(60)

Winner-Take-All Networks

161

(12)

Hamming networks

161

(2)

Maxnet

163

(1)

Simple competitive learning

164

(9)

Learning Vector Quantizers

173

(3)

Counterpropagation Networks

176

(4)

Adaptive Resonance Theory

180

(7)

Topologically Organized Networks

187

(14)

Self-organizing maps

188

(7)

Convergence*

195

(2)

Extensions

197

(4)

Distance-Based Learning

201

(3)

Maximum entropy

202

(1)

Neural gas

202

(2)

Neocognitron

204

(4)

Principal Component Analysis Networks

208

(5)

Conclusion

213

(1)

Exercises

214

(3)

Associative Models

217

(50)

Non-iterative Procedures for Association

219

(8)

Hopfield Networks

227

(17)

Discrete Hopfield networks

228

(9)

Storage capacity of Hopfield networks*

237

(4)

Continuous Hopfield networks

241

(3)

Brain-State-in-a-Box Network

244

(5)

Boltzmann Machines

249

(6)

Mean field annealing

254

(1)

Hetero-associators

255

(7)

Conclusion

262

(1)

Exercises

263

(4)

Optimization Methods

267

(40)

Optimization using Hopfield Networks

269

(10)

Traveling salesperson problem

270

(5)

Solving simultaneous linear equations

275

(1)

Allocating documents to multiprocessors

276

(3)

Iterated Gradient Descent

279

(1)

Simulated Annealing

280

(5)

Random Search

285

(2)

Evolutionary Computation

287

(13)

Evolutionary algorithms

288

(2)

Initialization

290

(1)

Termination criterion

290

(1)

Reproduction

291

(2)

Operators

293

(3)

Replacement

296

(3)

Schema Theorem*

299

(1)

Conclusion

300

(2)

Exercises

302

(5)

A Little Math

307

(8)

Calculus

307

(2)

Linear Algebra

309

(1)

Statistics

310

(5)

Data

315

(16)

Iris Data

315

(1)

Classification of Myoelectric Signals

316

(2)

Gold Prices

318

(1)

Clustering Animal Features

319

(1)

3-D Corners, Grid and Approximation

319

(4)

Eleven-City Traveling Salesperson Problem (Distances)

323

(1)

Daily Stock Prices of Three Companies, over the Same Period

324

(3)

Spiral Data

327

(4)

Bibliography

331

(8)

Index

339

What is included with this book?

The New copy of this book will include any supplemental materials advertised. Please check the title of the book to determine if it should include any access cards, study guides, lab manuals, CDs, etc.

The Used, Rental and eBook copies of this book are not guaranteed to include any supplemental materials. Typically, only the book itself is included. This is true even if the title states it includes any access cards, study guides, lab manuals, CDs, etc.

FREE SHIPPING

Amazon no longer offers textbook rentals. We do!

Elements of Artificial Neural Networks

Summary

Table of Contents

Supplemental Materials