상세정보

Exploring many aspects of neurocomputers, this book gives an overview of the network theory behind them, including a background review, basic concepts, associative networks, mapping networks, spatiotemporal networks, and adaptive resonance networks.

정보제공 :

CONTENTS
1 Introduction: What Is Neurocomputing? = 1
 1.1 Introduction = 1
  1.1.1 Overview of Neurocomputing = 1
 1.2 Neurocomputing as a Subject = 10
 1.3 The Relationship Between Neurocomputing and Neuroscience = 12
  1.3.1 Neurocomputing and Neuroscience = 12
  1.3.2 Hype = 13
 1.4 History of Neurocomputing = 14
  1.4.1 The Beginning = 14
  1.4.2 First Successes = 15
  1.4.3 The Quiet Years = 18
  1.4.4. Neurocomputing Takes Off = 18
 1.5 Guide to This Book = 19
2 Neural Network Concepts, Definitions, and Building Blocks = 21
 2.1 Neural Networks = 21
  2.1.1 Definition of a Neural Network = 22
 2.2 Connections = 26
  2.2.1 Connection Signal Data Types = 27
  2.2.2 Input Classes = 28
  2.2.3 Connection Geometries = 29
 2.3 Processing Elements = 37
  2.3.1 Transfer Functions and Local Memories = 37
  2.3.2 Slabs = 40
 2.4 N-Dimensional Geometry = 41
  2.4.1 Cubes = 41
  2.4.2 Spheres and Cubes = 42
3 Learning Laws: Self-Adaptation Equations = 46
 3.1 Definitions = 46
  3.1.1 Information Environments = 46
  3.1.2 Weight Space = 47
  3.1.3 Varieties of Training = 49
 3.2 Coincidence Learning = 51
  3.2.1 Hebb's Biological Learning Law = 51
  3.2.2 The Linear Associator = 52
  3.2.3 Hebb's Learning Law = 53
  3.2.4 The Pseudoinverse Formula = 55
 3.3 Performance Learning = 57
  3.3.1 The ADALINE = 58
  3.3.2 The Least Mean Squared Error Goal = 59
  3.3.3 The Widrow Learning Law = 60
 3.4 Competitive Learning = 64
  3.4.1 Kohonen's Layer = 66
  3.4.2 The Kohonen Learning Law = 67
  3.4.3 Estimation of the Probability Density Function = 68
 3.5 Filter Learning = 72
  3.5.1 The Flywheel Equation = 72
  3.5.2 The Instar = 74
  3.5.3 Grossberg's Learning Law = 74
 3.6 Spatiotemporal Learning = 76
  3.6.1 Temporal Sequences = 76
  3.6.2 The Kosko/Klopf Learning Law = 77
4 Associative Networks: Data Transformation Structures = 79
 4.1 Basic Definitions = 79
 4.2 Linear Associator Network = 81
 4.3 The Learnmatrix Network = 87
  4.3.1 Definition of the Learnmatrix Network = 88
  4.3.2 Learnmatrix Optical Anlysis = 89
  4.3.3 Learnmatrix Capacity = 91
 4.4 Recurrent Associative Networks = 95
  4.4.1 The Hopfield Network = 96
  4.4.2 The Brain State in a Box Network = 100
  4.4.3 Associative Network Theorems = 101
 4.5 Association Fascicles = 107
5 Mapping Networks: Muti-Layer Data Transformation Structures = 110
 5.1 The Mapping Implementation problem = 110
  5.1.1 Mapping Neural Networks = 111
  5.1.2 Measuring Function Approximation Accuracy = 111
  5.1.3 Training and Overtraining = 115
  5.1.4 Relationship to Statistical Regression = 120
 5.2 Kolmogorov's Theorem = 122
  5.2.1 Implications for Neurocomputing = 123
 5.3 The Backpropagation Neural Network = 124
  5.3.1 Architeture of the Backpropagation Network = 125
  5.3.2 Backpropagation Error Surfaces = 128
  5.3.3 Function Approximation with Backpropagation = 131
  5.3.4 Backpropagation Learning Laws = 133
 5.4 Self-Organizing Map = 138
  5.4.1 Architecture of the Self-Organizing Map Neural Network = 138
  5.4.2 Examples of the Operation of the Self-Organizing Map = 141
 5.5 Counterpropagation Network = 147
  5.5.1 Architecture of the Counterpropagation Neural Network = 147
  5.5.2 Variants of the Counterpropagation Network = 152
 5.6 Group Method of Data Handling = 155
  5.6.1 The GMDH Neural Network = 156
  5.6.2 GMDH Lessons = 162
6 Spatiotemporal, Stochastic, and Hierarchical Networks : Frontiers of Neurocomputing = 164
 6.1 Spatiotemporal Networks = 164
  6.1.1 Spatiotemporal Pattern Recognizer Neural Network = 166
  6.1.2 Recurrent Backpropagation Neural Network = 182
 6.2 Stochastic Networks = 192
  6.2.1 Finding Global Minima by Simulated Annealing = 192
  6.2.2 The Boltzmann Machine Network = 195
 6.3 Hierarchical Networks = 198
  6.3.1 Neocognitron Network = 198
  6.3.2 Combinatorial Hypercompression = 210
  6.3.3 Attention Mechanisms: Segmentation and Object Isolation = 214
7 Neurosoftware: Descriptions of Neural Network Structure = 220
 7.1 Neurosoftware: Coded Descriptions of Neural Network Structure = 221
 7.2 Software Interfaces Between Computers and Neurocomputers = 222
 7.3 AXON Language = 226
  7.3.1 AXON Structure =  226
  7.3.2 Parameter Definition Block = 229
  7.3.3 Processing Element and Slab Definition Block = 231
  7.3.4 Network Creation and Connection Definition Block = 233
  7.3.5 Scheduling Function Block = 238
  7.3.6 Function Definition Block = 239
 7.4 AXON Examples = 242
  7.4.1 Backpropagation = 242
  7.4.2 Counterpropagation = 248
8 Neurocomputers: Machines for Implementing Neural Networks = 259
 8.1 Neurocomputer Fundamentals = 260
  8.1.1 Neurocomputers as Computer Coprocessors = 260
  8.1.2 Performance Measures = 263
  8.1.3 Taxonomy = 266
 8.2 Analog and Hybrid Neurocomputer Design Fundamentals = 272
  8.2.1 Overview = 272
  8.2.2 Primacy of Input Processing and Weight Modification = 276
  8.2.3 Transfer Function Implementation = 284
 8.3 Analog and Hybrid Neurocomputer Design Examples = 287
  8.3.1 Electro-Optic Neurocomputer Design Examples = 287
  8.3.2 Optical Neurocomputer = 289
  8.3.3 MNOS/CCD Electronic Neurocomputer Chip = 292
 8.4 Digital Nerocomputer Design Fundamentals = 297
  8.4.1 Overview = 297
  8.4.2 Fully Implemented Design Approaches = 298
  8.4.3 Virtual Desing Approaches = 301
 8.5 Digital Neurocomputer Design Examples = 305
  8.5.1 MarkⅢ Neurocomputer = 305
  8.5.2 MarkⅣ Neurocomputer = 307
9 Neurocomputing Applications: Sensor Processing, Control, and Data Anlysis = 317
 9.1 Neurocomputing Applications Engineering = 317
  9.1.1 Solving Problems with Neurocomputing = 318
  9.1.2 Functional Specification Development = 320
 9.2 Sensor Processing = 323
  9.2.1 Character Recognizer = 323
  9.2.2 Cottrell/Munro/Zipser Technique = 325
  9.2.3 Noise Removal from Time-Series Signals = 337
 9.3 Control = 342
  9.3.1 Vision-Based Broomstick Balancer = 342
  9.3.2 Automobile Autopilot = 345
 9.4 Data Analysis = 349
  9.4.1 Loan Application Scoring = 349
  9.4.2 NETtalk = 351
  9.4.3 The Instant Physician = 354
A  Neurocomputing Projects : Developing New Capabilities that Succeed in the Marketplace = 358
 A.1 Business Plan Development = 359
  A.1.1 Project Definition = 361
  A.1.2 Defining Goals = 362
  A.1.3 Technical Feasibility = 363
  A.1.4 Market Analysis = 363
  A.1.5 Development Plan = 368
  A.1.6 Marketing and Sales Plan = 371
  A.1.7 Production Plan = 374
  A.1.8 Organization and Personnel = 376
  A.1.9 Schedule = 377
  A.1.10 Budget = 378
  A.1.11 Financing and Ownership = 383
 A.2 Writing a Proposal = 388
  A.2.1 RFPs and RFQs = 389
  A.2.2 Proposal Organization = 389
  A.2.3 Proposal Writing = 392
 A.3 Planning and Managing Development = 394
  A.3.1 The Development Planning Process = 394
  A.3.2 Project Management = 402
INDEX = 422