Table of Contents
Preface v
About the Author vii
Acknowledgements ix
1 Introduction 1
1.1 Introduction 1
1.2 Time Domain Data 2
1.2.1 Average 3
1.2.2 Variance 3
1.2.3 Kurtosis 5
1.3 Frequency Domain 5
1.4 Time-Frequency Domain 8
1.5 Fractals 9
1.6 Stationarity 12
1.7 Common Mistakes on Handling Data 15
1.8 Outline of the Book 17
1.9 Conclusions 19
References 19
2 Multi-layer Perceptron 23
2.1 Introduction 23
2.2 Multi-layer Perceptron 24
2.3 Training the Multi-layered Perccptron 26
2.4 Back-propagation Method 27
2.5 Scaled Conjugate Method 29
2.6 Multi-layer Perceptron Classifier 32
2.7 Applications to Economic Modelling 33
2.8 Application to a Steam Generator 35
2.9 Application to Cylindrical Shells 36
2.10 Application to Interstate Conflict 37
2.11 Conclusions 39
References 39
3 Radial Basis Function 43
3.1 Introduction 43
3.2 Radial Basis Function 44
3.3 Model Selection 48
3.4 Application to Interstate Conflict 49
3.5 Call Behaviour Classification 50
3.6 Modelling the CPI 53
3.7 Modelling Steam Generator 54
3.8 Conclusions 56
References 56
4 Automatic Relevance Determination 59
4.1 Introduction 59
4.2 Mathematical Basis of the Automatic Relevance Determination 60
4.2.1 Neural networks 60
4.2.2 Bayesian framework 63
4.2.3 Automatic relevance determination 66
4.3 Application to Interstate Conflict 68
4.4 Applications of ARD in Inflation Modelling 70
4.5 Conclusions 72
References 72
5 Bayesian Networks 77
5.1 Introduction 77
5.2 Neural Networks 80
5.3 Hybrid Monte Carlo 81
5.4 Shadow Hybrid Monte Carlo (SHMC) Method 84
5.5 Separable Shadow Hybrid Monte Carlo 89
5.6 Comparison of Sampling Methods 91
5.7 Interstate Conflict 92
5.8 Conclusions 93
References 93
6 Support Vector Machines 97
6.1 Introduction 97
6.2 Support Vector Machines for Classification 98
6.3 Support Vector Regression 103
6.4 Conflict Modelling 106
6.5 Steam Generator 107
6.6 Conclusions 108
References 108
7 Fuzzy Logic 113
7.1 Introduction 113
7.2 Fuzzy Logic Theory 115
7.3 Neuro-fuzzy Models 118
7.4 Steam Generator 122
7.5 Interstate Conflict 123
7.6 Conclusions 126
References 126
8 Rough Sets 133
8.1 Introduction 133
8.2 Rough Sets 134
8.2.1 Information system 135
8.2.2 The indiscernibility relation 135
8.2.3 Information table and data representation 136
8.2.4 Decision rules induction 136
8.2.5 The lower and upper approximation of sets 137
8.2.6 Set approximation 138
8.2.7 The reduct 139
8.2.8 Boundary region 139
8.2.9 Rough membership functions 140
8.3 Discretization Methods 140
8.3.1 Equal-width-bin (EWB) partitioning 140
8.3.2 Equal-frequency-bin (EFB) partitioning 141
8.4 Rough Set Formulation 141
8.5 Rough Sets vs. Fuzzy Sets 142
8.6 Multi-layer Perceptron Model 143
8.7 Neuro-rough Model 144
8.7.1 Bayesian training on rough sets 144
8.7.2 Markov Chain Monte Carlo (MCMC) 146
8.8 Modelling of HIV 147
8.9 Application to Modelling the Stock Market 150
8.10 Interstate Conflict 153
8.11 Conclusions 154
References 154
9 Hybrid Machines 159
9.1 Introduction 159
9.2 Hybrid Machine 160
9.2.1 Bayes optimal classifier 161
9.2.2 Bayesian model averaging 162
9.2.3 Bagging 162
9.2.4 Boosting 163
9.2.5 Stacking 163
9.2.6 Evolutionary machines 164
9.3 Theory of Hybrid Networks 164
9.3.1 Equal weights 166
9.3.2 Variable weights 167
9.4 Condition Monitoring 168
9.5 Caller Behaviour 171
9.6 Conclusions 173
References 174
10 Auto-associative Networks 179
10.1 Introduction 179
10.2 Auto-associative Networks 180
10.3 Principal Component Analysis 181
10.4 Missing Data Estimation 183
10.5 Genetic Algorithm (GA) 184
10.6 Machine Learning 187
10.7 Modelling HIV 188
10.8 Artificial Beer Taster 190
10.9 Conclusions 192
References 192
11 Evolving Networks 199
11.1 Introduction 199
11.2 Machine Learning 201
11.3 Genetic Algorithm 202
11.4 Learn++ Method 203
11.5 Incremental Learning Method Using Genetic Algorithm (ILUGA) 206
11.6 Optical Character Recognition (OCR) 208
11.7 Wine Recognition 210
11.8 Financial Analysis 212
11.9 Condition Monitoring of Transformers 214
11.10 Conclusions 216
References 216
12 Causality 221
12.1 Introduction 221
12.2 Correlation 222
12.3 Causality 223
12.4 Theories of Causality 224
12.4.1 Transmission theory of causality 225
12.4.2 Probability theory of causality 225
12.4.3 Projectile theory of causality 226
12.4.4 Causal calculus and structural learning 226
12.4.5 Granger causality 227
12.4.6 Structural learning 227
12.4.7 Manipulation theory 227
12.4.8 Process theory 228
12.4.9 Counterfactual theory 228
12.4.10 Neyman-Rubin causal model 229
12.4.11 Causal calculus 230
12.4.12 Inductive causation (IC) 235
12.5 How to Detect Causation? 236
12.6 Causality and Artificial Intelligence 237
12.7 Causality and Rational Decision 238
12.8 Conclusions 238
References 239
13 Gaussian Mixture Models 245
13.1 Introduction 245
13.2 Gaussian Mixture Models 246
13.3 EM Algorithm 249
13.4 Condition Monitoring: Transformer Bushings 251
13.5 Condition Monitoring: Cylindrical Shells 252
13.6 Condition Monitoring: Bearings 253
13.7 Conclusions 257
References 258
14 Hidden Markov Models 263
14.1 Introduction 263
14.2 Hidden Markov Models 264
14.3 Condition Monitoring: Motor Bearing Faults 269
14.4 Speaker Recognition 274
14.5 Conclusions 277
References 277
15 Reinforcement Learning 283
15.1 Introduction 283
15.2 Reinforcement Learning: TD-Lambda 284
15.3 Game Theory 287
15.4 Multi-agent Systems 288
15.5 Modelling the Game of Lerpa 289
15.6 Modelling of Tic-Tac-Toe 295
15.7 Conclusions 300
References 300
16 Conclusion Remarks 305
16.1 Summary of the Book 305
16.2 Implications of Artificial Intelligence 307
References 309
Index 311
