Artificial neural systems : principle and practice

An intelligent system is one which exhibits characteristics including, but not limited to, learning, adaptation, and problem-solving. Artificial Neural Network (ANN) Systems are intelligent systems designed on the basis of statistical models of learning that mimic biological systems such as the human central nervous system. Such ANN systems represent the theme of this book. This book also describes concepts related to evolutionary methods, clustering algorithms, and others networks which are complementary to ANN system.The book is divided into two parts. The first part explains basic concepts derived from the natural biological neuron and introduces purely scientific frameworks used to develop a viable ANN model. The second part expands over to the design, analysis, performance assessment, and testing of ANN models. Concepts such as Bayesian networks, multi-classifiers, and neuromorphic ANN systems are explained, among others.Artificial Neural Systems: Principles and Practice takes a developmental perspective on the subject of ANN systems, making it a beneficial resource for students undertaking graduate courses and research projects, and working professionals (engineers, software developers) in the field of intelligent systems design. CONTENTS......Page 5 FOREWORD......Page 9 PREFACE......Page 11 Principles......Page 13 A BIOLOGICAL NEURON......Page 14 TRANSMISSION ACROSS SYNAPSES......Page 16 AN ARTIFICIAL NEURON......Page 18 ACKNOWLEDGEMENTS......Page 23 REFERENCES......Page 24 INTEGRATE-AND-FIRE NEURON......Page 25 PROBABILITY......Page 27 STEIN MODEL OF NEURON......Page 35 REFERENCES......Page 36 A FUZZY NEURON......Page 38 The Fuzzy-logic Neuron......Page 40 PRINCIPLES OF ARTIFICIAL NEURAL NETWORK ANALYSIS AND DESIGN......Page 42 The Wave Neural Networks......Page 46 CONFLICT OF INTEREST......Page 48 REFERENCES......Page 49 INTRODUCTION......Page 50 DENSITY BASED ALGORITHMS: CLUSTERING ALGORITHMS......Page 51 NATURE-BASED ALGORITHMS......Page 54 Evolutionary Algorithm and Programming......Page 55 Genetic Algorithm......Page 57 GA Operators......Page 58 NETWORK METHOD: EDGES AND NODES......Page 59 MULTI-LAYERED PERCEPTRON......Page 61 REAL-TIME APPLICATIONS OF STATE-OF-THE-ART ANN SYSTEMS......Page 63 Intelligence......Page 64 An Artificial Neural Network (ANN) system......Page 65 Receiver’s Operating Characteristics (ROC)......Page 70 Hypothesis Testing......Page 72 Chi-squared (Goodness-of-fit) Test......Page 75 REFERENCES......Page 77 INTRODUCTION......Page 79 QUANTUM LOGIC AND QUANTUM MATHEMATICS......Page 80 Quantum Gates (Primitives)......Page 81 Quantum Algebra......Page 86 QUANTUM NEURAL NETWORK......Page 88 Memristance......Page 90 HODGKIN-HUXLEY NEURON......Page 94 REFERENCES......Page 95 Practices......Page 97 INTRODUCTION......Page 98 THE ADAPTIVE LINEAR NEURON (ADALINE)......Page 99 THE RECURSIVE-LEAST-SQUARE (RLS) ALGORITHM......Page 103 MULTI-AGENT NETWORK......Page 105 NEUROMORPHIC NETWORK......Page 109 Gaussian Mixture Model......Page 113 K-means......Page 117 Radial Basis Function (RBF)......Page 118 Generative Topographic Mapping (GTM)......Page 121 NEURO-FUZZY SYSTEM......Page 123 RESEARCH AND APPLICATIONS OF ANN SYSTEMS......Page 125 REFERENCES......Page 126 INTRODUCTION......Page 128 Probabilistic Convergent Network (PCN)......Page 129 PCN Network Architecture......Page 131 Recognition or Classification......Page 132 THE EPCN......Page 133 The EPCN Software Implementation......Page 136 Multi-Layer Perceptron (MLP)......Page 137 Mixture Density Network (MDN)......Page 141 Helmholtz Machine......Page 148 Introduction: Chi-Squared Probability Density Function......Page 154 The Dynamics......Page 156 Fusion......Page 158 Generalized Likelihood Ratio Test (GLRT)......Page 159 Wald Test......Page 160 Wald Test Procedure:......Page 161 REFERENCES......Page 162 INTRODUCTION......Page 164 FACTORIAL SELECTION......Page 165 Comparison to Other Similar Coding Scheme for Multi-class Problems......Page 174 THE GROUP METHOD OF SELECTION......Page 175 Applications of GMDH......Page 178 REFERENCES......Page 179 INTRODUCTION......Page 181 RANDOM-NUMBER GENERATORS......Page 182 MARKOV CHAIN......Page 183 HYBRID MARKOV CHAIN (HMC)......Page 184 Momentum Heat-Bath......Page 185 Molecular Dynamics......Page 186 Acceptance Criteria......Page 187 1. Verlet Integrator......Page 188 2. Velocity Verlet......Page 190 Gibbs Sampling......Page 191 The Restricted Boltzmann Machine (RBM)......Page 192 Energy Dynamics and Learning......Page 195 A DEEP BELIEF NETWORK OF BOLTZMANN MACHINES......Page 196 Boltzmann Machine Learning Algorithm......Page 199 The Partition Function: Annealed Importance Sampling (AIS)......Page 200 Pre-Training of Deep Belief Network......Page 201 Dynamic Biases of a DBN......Page 203 REFERENCES......Page 207 INTRODUCTION......Page 208 MEMRISTIC NEURAL NETWORKS......Page 209 QUANTUM EXPERT SYSTEMS......Page 213 Initialization......Page 214 Behaviour......Page 215 Learning Algorithm......Page 220 DEEP BELIEF NETWORKS (DBN) IN INDUSTRY......Page 222 REFERENCES......Page 225 INTRODUCTION......Page 227 EXTENSION OF HYBRID MONTE CARLO......Page 228 NEUROMORPHIC NETWORKS II......Page 237 CONCLUSION......Page 242 REFERENCES......Page 244 SUBJECT INDEX......Page 245 An intelligent system is one which exhibits characteristics including, but not limited to, learning, adaptation, and problem-solving. Artificial Neural Network (ANN) Systems are intelligent systems designed on the basis of statistical models of learning that mimic biological systems such as the human central nervous system. Such ANN systems represent the theme of this book. This book also describes concepts related to evolutionary methods, clustering algorithms, and other networks which are complementary to ANN systems.The book is divided into two parts. The first part explains basic concepts derived from the natural biological neuron and introduces purely scientific frameworks used to develop a viable ANN model. The second part expands over to the design, analysis, performance assessment, and testing of ANN models. Concepts such as Bayesian networks, multi-classifiers, and neuromorphic ANN systems are explained, among others.Artificial Neural Systems: Principles and Practice takes a developmental perspective on the subject of ANN systems, making it a beneficial resource for students undertaking graduate courses and research projects, and working professionals (engineers, software developers) in the field of intelligent systems design.