Foreign-Exchange-Rate Forecasting with Artificial Neural Network

Table of Contents
Preface .......................................................................................................xi
Biographies of Three Authors of the Book............................................xv
List of Figures ........................................................................................xvii
List of Tables ...........................................................................................xxi
Part I: Forecasting Foreign Exchange Rates with Artificial
Neural Networks: An Analytical Survey............................1
1 Are Foreign Exchange Rates Predictable? — A Literature
Review from Artificial Neural Networks Perspective .........................3
1.1 Introduction .................................................................................................... 3
1.2 Literature Collection....................................................................................... 5
1.3.1 Basic Classifications and Factors Summarization ................................ 7
1.3.2 Factor Analysis...................................................................................... 8
1.4 Implications and Research Topics................................................................ 21
1.5 Conclusions .................................................................................................. 23
Part II: Basic Learning Principles of Artificial Neural
2 Basic Learning Principles of Artificial Neural Networks..................27
2.1 Introduction .................................................................................................. 27
2.2 Basic Structure of the BPNN Model ............................................................ 28
2.3 Learning Process of the BPNN Algorithm................................................... 30
2.4 Weight Update Formulae of the BPNN Algorithm...................................... 31
2.5 Conclusions .................................................................................................. 37
3 Data Preparation in Neural Network Data Analysis .........................39
3.1 Introduction .................................................................................................. 39
3.2 Neural Network for Data Analysis............................................................... 42
1.3 Analytical Results and Factor Investigation .................................................. 7
Networks and Data Preparation .....................................25
3.3 An Integrated Data Preparation Scheme ...................................................... 44
3.3.1 Integrated Data Preparation Scheme for Neural Network
Data Analysis ...................................................................................... 44
3.3.2 Data Pre-Analysis Phase ..................................................................... 46
3.3.3 Data Preprocessing Phase.................................................................... 51
3.3.4 Data Post-Analysis Phase.................................................................... 56
3.4 Costs–Benefits Analysis of the Integrated Scheme ..................................... 59
3.5 Conclusions .................................................................................................. 61
Part III: Individual Neural Network Models with Optimal
Learning Rates and Adaptive Momentum
Factors for Foreign Exchange Rates Prediction ..........63
4 Forecasting Foreign Exchange Rates Using
an Adaptive Back-Propagation Algorithm with Optimal
4.1 Introduction .................................................................................................. 65
4.2 BP Algorithm with Optimal Learning Rates
and Momentum Factors................................................................................ 68
4.2.1 Optimal Learning Rates Determination .............................................. 68
4.2.2 Determination of Optimal Momentum Factors................................... 76
4.3.1 Data Description and Experiment Design........................................... 78
4.3.2 Experimental Results........................................................................... 80
4.4 Concluding Remarks .................................................................................... 84
5 An Online BP Learning Algorithm with Adaptive Forgetting
Factors for Foreign Exchange Rates Forecasting..............................87
5.1 Introduction .................................................................................................. 87
5.2 An Online BP Learning Algorithm with Adaptive Forgetting Factors........ 88
5.3 Experimental Analysis ................................................................................. 94
5.3.1 Data Description and Experiment Design........................................... 94
5.3.2 Experimental Results........................................................................... 96
5.4 Conclusions .................................................................................................. 99
6 An Improved BP Algorithm with Adaptive Smoothing
Momentum Terms for Foreign Exchange Rates Prediction ...........101
6.1 Introduction ................................................................................................ 101
Learning Rates and Momentum Factors ...........................................65
4.3 Experiment Study......................................................................................... 78
6.2 Formulation of the Improved BP Algorithm.............................................. 103
6.2.1 Determination of Adaptive Smoothing Momentum ......................... 103
6.2.2 Formulation of the Improved BPNN Algorithm............................... 106
6.3 Empirical Study.......................................................................................... 108
6.3.1 Data Description and Experiment Design......................................... 109
6.3.2 Forecasting Results and Comparisons .............................................. 109
6.3.3 Comparisons of Different Learning Rates ........................................ 112
6.3.4 Comparisons with Different Momentum Factors ............................ 113
6.3.5 Comparisons with Different Error Propagation Methods ................. 114
6.3.6 Comparisons with Different Numbers of Hidden Neurons............... 115
6.3.7 Comparisons with Different Hidden Activation Functions .............. 116
6.4 Comparisons of Three Single Neural Network Models............................. 117
6.5 Conclusions ................................................................................................ 117
Part IV: Hybridizing ANN with Other Forecasting
Forecasting.....................................................................119
7 Hybridizing BPNN and Exponential Smoothing for Foreign
8 A Nonlinear Combined Model Hybridizing ANN and GLAR
for Exchange Rates Forecasting ........................................................133
8.1 Introduction ................................................................................................ 133
8.2 Model Building Processes.......................................................................... 136
8.2.1 Generalized Linear Auto-Regression (GLAR) Model...................... 136
8.2.2 Artificial Neural Network (ANN) Model ......................................... 138
8.2.3 A Hybrid Model Integrating GLAR with ANN................................ 139
8.2.4 Combined Forecasting Models.......................................................... 141
Techniques for Foreign Exchange Rates
Exchange Rate Prediction .................................................................121
7.1 Introduction ................................................................................................ 121
7.2 Basic Backgrounds ..................................................................................... 123
7.2.1 Exponential Smoothing Forecasting Model ...................................... 123
7.2.2 Neural Network Forecasting Model .................................................. 125
7.3 A Hybrid Model Integrating BPNN and Exponential Smoothing ............. 127
7.4 Experiments ................................................................................................ 129
7.5 Conclusions ................................................................................................ 130
8.2.5 A Nonlinear Combined (NC) Forecasting Model............................. 142
8.2.6 Forecasting Evaluation Criteria......................................................... 145
8.3 Empirical Analysis ..................................................................................... 148
8.3.1 Data Description................................................................................ 148
8.3.2 Empirical Results .............................................................................. 148
8.4 Conclusions ................................................................................................ 153
Market Tendency Exploration ..........................................................155
9.1 Introduction ................................................................................................ 155
9.2 Formulation of the Hybrid GA-SVM Model ............................................. 158
9.2.1 Basic Theory of SVM ....................................................................... 158
9.2.2 Feature Selection with GA for SVM Modeling ................................ 160
9.2.3 A Hybrid GASVM Model................................................................. 164
9.3 Empirical Study.......................................................................................... 165
9.3.1 Research Data.................................................................................... 165
9.3.2 Descriptions of Other Comparable Forecasting Models................... 167
9.3.3 Experiment Results ........................................................................... 168
9.4 Comparisons of Three Hybrid Neural Network Models............................ 172
9.5 Conclusions ................................................................................................ 173
Part V: Neural Network Ensemble for Foreign Exchange
Rates Forecasting............................................................175
10 Forecasting Foreign Exchange Rates with a Multistage Neural
Network Ensemble Model................................................................177
10.1 Introduction ............................................................................................ 177
10.2 Motivations for Neural Network Ensemble Model................................ 179
10.3 Formulation of Neural Network Ensemble Model................................. 181
10.3.1 Framework of Multistage Neural Ensemble Model.................... 181
10.3.2 Preprocessing Original Data........................................................ 182
10.3.3 Generating Individual Neural Predictors .................................... 185
10.3.4 Selecting Appropriate Ensemble Members................................. 187
10.3.5 Ensembling the Selecting Members............................................ 192
10.4 Empirical Analysis ................................................................................. 196
10.4.1 Experimental Data and Evaluation Criterion .............................. 196
10.4.2 Experiment Design...................................................................... 196
9 A Hybrid GA-Based SVM Model for Foreign Exchange
10.4.3 Experiment Results and Comparisons ........................................ 198
10.5 Conclusions ............................................................................................ 201
11 Neural Networks Meta-Learning for Foreign Exchange Rate
Ensemble Forecasting.......................................................................203
11.1 Introduction ............................................................................................ 203
11.2 Introduction of Neural Network Learning Paradigm ............................. 204
11.3 Neural Network Meta-Learning Process for Ensemble ......................... 206
11.3.1 Basic Background of Meta-Learning .......................................... 206
11.3.2 Data Sampling ............................................................................. 207
11.3.3 Individual Neural Network Base Model Creation ...................... 209
11.3.4 Neural Network Base Model Pruning ......................................... 210
11.3.5 Neural-Network-Based Meta-Model Generation........................ 212
11.4 Empirical Study...................................................................................... 213
11.4.1 Research Data and Experiment Design....................................... 213
11.4.2 Experiment Results ..................................................................... 215
11.5 Conclusions ............................................................................................ 216
12 Predicting Foreign Exchange Market Movement Direction
Using a Confidence-Based Neural Network Ensemble Model...... 217
12.1 Introduction ............................................................................................ 217
12.2 Formulation of Neural Network Ensemble Model................................. 219
12.2.1 Partitioning Original Data Set..................................................... 220
12.2.2 Creating Individual Neural Network Classifiers......................... 221
12.2.3 BP Network Learning and Confidence Value Generation.......... 222
12.2.4 Confidence Value Transformation.............................................. 223
12.2.5 Integrating Multiple Classifiers into an Ensemble Output.......... 223
12.3 Empirical Study...................................................................................... 226
12.4 Comparisons of Three Ensemble Neural Networks............................... 230
12.5 Conclusions ............................................................................................ 230
13 Foreign Exchange Rates Forecasting with Multiple Candidate
Models: Selecting or Combining? A Further Discussion ..............233
13.1 Introduction ............................................................................................ 233
13.2 Two Dilemmas and Their Solutions....................................................... 237
13.3 Empirical Analysis ................................................................................. 242
13.4 Conclusions and Future Directions ........................................................ 244
Part VI: Developing an Intelligent Foreign Exchange
Rates Forecasting and Trading Decision
Support System..............................................................247
14 Developing an Intelligent Forex Rolling Forecasting
and Trading Decision Support System I: Conceptual
Framework, Modeling Techniques and System
Implementations ............................................................................... 249
14.1 Introduction ............................................................................................ 249
14.2 System Framework and Main Functions................................................ 250
14.3 Modeling Approach and Quantitative Measurements............................ 252
14.3.1 BPNN-Based Forex Rolling Forecasting Sub-System................ 253
14.3.2 Web-Based Forex Trading Decision Support System ................ 263
14.4 Development and Implementation of FRFTDSS................................... 269
14.4.1 Development of the FRFTDSS ................................................... 269
14.4.2 Implementation of the FRFTDSS ............................................... 270
14.5 Conclusions ............................................................................................ 274
15 Developing an Intelligent Forex Rolling Forecasting
and Trading Decision Support System II: An Empirical
and Comprehensive Assessment......................................................275
15.1 Introduction ............................................................................................ 275
15.2 Empirical Assessment on Performance of FRFTDSS ........................... 276
15.2.1 Parametric Evaluation Methods .................................................. 276
15.2.2 Nonparametric Evaluation Methods ........................................... 278
15.3 Performance Comparisons with Classical Models................................. 280
15.3.1 Selection for Comparable Classical Models ............................... 280
15.3.2 Performance Comparison Results with Classical Models .......... 280
15.4 Performance Comparisons with Other Systems..................................... 281
15.4.1 Searching for Existing Forex Forecasting Systems .................... 281
15.4.2 Performance Comparisons with Other Existing Systems ........... 283
15.4.3 A Comprehensive Comparison Analysis .................................... 285
15.5 Discussions and Conclusions ................................................................. 288
References............................................................................................... 291
Subject Index.......................................................................................... 311

Another Neural Network Boks
Another Time Series Books
Download