Chen / Abraham Tree-Structure based Hybrid Computational Intelligence

Theoretical Foundations and Applications
1. Auflage 2009
ISBN: 978-3-642-04739-8
Verlag: Springer
Format: PDF
Kopierschutz: 1 - PDF Watermark

E-Book, Englisch, Band 2, 206 Seiten

Reihe: Intelligent Systems Reference Library

Research in computational intelligence is directed toward building thinking machines and improving our understanding of intelligence. As evident, the ultimate achievement in this field would be to mimic or exceed human cognitive capabilities including reasoning, recognition, creativity, emotions, understanding, learning and so on. In this book, the authors illustrate an hybrid computational intelligence framework and it applications for various problem solving tasks. Based on tree-structure based encoding and the specific function operators, the models can be flexibly constructed and evolved by using simple computational intelligence techniques. The main idea behind this model is the flexible neural tree, which is very adaptive, accurate and efficient. Based on the pre-defined instruction/operator sets, a flexible neural tree model can be created and evolved.

This volume comprises of 6 chapters including an introductory chapter giving the fundamental definitions and the last Chapter provides some important research challenges. Academics, scientists as well as engineers engaged in research, development and application of computational intelligence techniques and data mining will find the comprehensive coverage of this book invaluable.

Autoren/Hrsg.

Chen, Yuehui

Abraham, Ajith

Inhaltsverzeichnis

1;Preface;7
2;Contents;10
3;Part I Foundations of Computational Intelligence;14
3.1;1 Foundations of Computational Intelligence;15
3.1.1;1.1 Introduction;15
3.1.2;1.2 Evolutionary Algorithms;15
3.1.2.1;1.2.1 Genetic Programming;19
3.1.2.2;1.2.2 Estimation of Distribution Algorithm;23
3.1.2.3;1.2.3 Population-Based Incremental Learning;25
3.1.2.4;1.2.4 Probabilistic Incremental Program Evolution;26
3.1.3;1.3 Swarm Intelligence;30
3.1.3.1;1.3.1 Particle Swarm Optimization;31
3.1.3.2;1.3.2 Ant Colony Optimization;33
3.1.4;1.4 Artificial Neural Networks;35
3.1.4.1;1.4.1 Architecture and Learning Algorithm;36
3.1.4.2;1.4.2 Multilayer Perceptron;38
3.1.4.3;1.4.3 Back-Propagation Algorithm;38
3.1.4.4;1.4.4 Evolutionary Algorithm Based Training;40
3.1.4.5;1.4.5 Self Organizing Feature Maps;41
3.1.4.6;1.4.6 Radial Basis Function;42
3.1.4.7;1.4.7 Recurrent Neural Networks;42
3.1.4.8;1.4.8 Adaptive Resonance Theory;43
3.1.5;1.5 Fuzzy Systems;43
3.1.5.1;1.5.1 The Definition of Fuzzy Sets;43
3.1.6;1.6 Takagi-Sugeno Fuzzy Model;44
3.1.6.1;1.6.1 Universal Approximation Property;45
3.1.6.2;1.6.2 Fuzzy Expert Systems - Design Challenges;45
3.1.7;1.7 Probabilistic Computing;46
3.1.8;1.8 Hybrid Intelligent Systems;47
3.1.9;1.9 Models of Hybrid Intelligent Systems;48
4;Part II Flexible Neural Trees;49
4.1;2 Flexible Neural Tree: Foundations and Applications;50
4.1.1;2.1 Introduction to Flexible Neural Tree;50
4.1.2;2.2 Flexible Neural Tree Algorithms;51
4.1.2.1;2.2.1 Encoding and Evaluation;51
4.1.2.2;2.2.2 Flexible Neuron Instructor;51
4.1.2.3;2.2.3 Fitness Function;53
4.1.2.4;2.2.4 Structure and Parameter Learning;53
4.1.2.5;2.2.5 Flexible Neural Tree Applications;55
4.1.2.6;2.2.6 Exchange Rate Forecasting;75
4.1.2.7;2.2.7 Face Recognition;80
4.1.2.8;2.2.8 Microarray-Based Cancer Classification;84
4.1.2.9;2.2.9 Protein Fold Recognition;87
4.1.3;2.3 Multi Input Multi Output Flexible Neural Tree;90
4.1.4;2.4 Representation and Calculation of the MIMO FNT;91
4.1.4.1;2.4.1 Hybrid Algorithm for Structure and Parameter Learning;93
4.1.4.2;2.4.2 Hybrid Algorithm for Flexible Neural Tree Model;95
4.1.4.3;2.4.3 Illustrative Examples;95
4.1.5;2.5 Ensemble of Flexible Neural Tree;100
4.1.5.1;2.5.1 The Basic Ensemble Method;101
4.1.5.2;2.5.2 The Generalized Ensemble Method;101
4.1.5.3;2.5.3 The LWPR Method;101
4.1.5.4;2.5.4 Stock Index Forecasting Problem;102
4.1.6;2.6 Stock Index Forecasting Experimental Illustrations;104
5;Part III Hierarchical Neural Networks;108
5.1;3 Hierarchical Neural Networks;109
5.1.1;3.1 Hierarchical Radial Basis Function Neural Networks;109
5.1.1.1;3.1.1 The Radial Basis Function Network;110
5.1.1.2;3.1.2 Automatic Design of Hierarchical Radial Basis Function;111
5.1.1.3;3.1.3 Tree Structure Optimization by Extended Compact;112
5.1.1.4;3.1.4 Parameter Optimization Using Differential Evolution;112
5.1.1.5;3.1.5 Procedure of The General Learning Algorithm;113
5.1.1.6;3.1.6 Variable Selection in the HRBF Network Paradigms;113
5.1.1.7;3.1.7 Experimental Illustrations;114
5.1.1.8;3.1.8 Face Recognition;115
5.1.2;3.2 Hierarchical B-Spline Neural Networks;118
5.1.2.1;3.2.1 The B-Spline Network;118
5.1.3;3.3 Automatic Design of HB-Spline Network;119
5.1.3.1;3.3.1 Encode and Calculation for HB-Spline;119
5.1.3.2;3.3.2 Tree Structure and Parameter Optimization;120
5.1.3.3;3.3.3 Procedure of the General Learning Algorithm;121
5.1.3.4;3.3.4 Variable Selection in the Hierarchical B-Spline Network;121
5.1.3.5;3.3.5 Experimental Illustrations;121
5.1.3.6;3.3.6 Wisconsin Breast Cancer Detection;121
5.1.3.7;3.3.7 Time-Series Forecasting;123
5.1.4;3.4 Hierarchical Wavelet Neural Networks;128
5.1.4.1;3.4.1 Wavelet Neural Network;128
5.1.5;3.5 Automatic Design of Hierarchical Wavelet Neural Network;129
5.1.5.1;3.5.1 Ant Programming for Evolving the Architecture of;129
5.1.5.2;3.5.2 Parameter Optimization Using Differential Evolution;131
5.1.5.3;3.5.3 Procedure of the General Learning Algorithm for HWNN;131
5.1.5.4;3.5.4 Variable Selection Using HWNN Paradigms;132
5.1.5.5;3.5.5 Experimental Illustrations;132
5.1.5.6;3.5.6 Application to Jenkins-Box Time-Series;134
6;Part IV Hierarchical Fuzzy Systems;136
6.1;4 Hierarchical Fuzzy Systems;137
6.1.1;4.1 Introduction;137
6.1.2;4.2 Takagi-Sugeno Fuzzy Inference System (TS-FS);139
6.1.3;4.3 Hierarchical TS-FS: Encoding and Evaluation;139
6.1.3.1;4.3.1 Encoding;140
6.1.3.2;4.3.2 Evaluation;141
6.1.3.3;4.3.3 Objective Function;142
6.1.4;4.4 Evolutionary Design of Hierarchical TS-FS;143
6.1.4.1;4.4.1 Algorithm for Designing Hierarchical TS-FS Model;143
6.1.4.2;4.4.2 Feature/Input Selection with Hierarchical TS-FS;144
6.1.5;4.5 Experimental Illustrations;145
6.1.5.1;4.5.1 Systems Identification;146
6.1.5.2;4.5.2 Chaotic Time-Series of Mackey-Glass;147
6.1.5.3;4.5.3 Iris Data Classification;150
6.1.5.4;4.5.4 Wine Data Classification;152
7;Part V Reverse Engineering of Dynamical Systems;156
7.1;5 Reverse Engineering of Dynamic Systems;157
7.1.1;5.1 Introduction;157
7.1.2;5.2 Calculation and Representation of Additive Models;158
7.1.3;5.3 Hybrid Algorithm;159
7.1.3.1;5.3.1 Tree-Structure Based Evolutionary Algorithm;159
7.1.3.2;5.3.2 Evolving an Optimal or Near-Optimal Structure of;160
7.1.3.3;5.3.3 Parameter Optimization;162
7.1.3.4;5.3.4 Summary of General Learning Algorithm;164
7.1.3.5;5.3.5 Experimental Illustrations;165
7.1.3.6;5.3.6 Discussions;171
7.1.4;5.4 Inferring a System of Differential Equations;174
7.1.5;5.5 Inference of Differential Equation Models by Multi Expression Programming;175
7.1.5.1;5.5.1 Structure Optimization by the MEP;175
7.1.5.2;5.5.2 Parameter Optimization by Particle Swarm Optimization;176
7.1.5.3;5.5.3 Fitness Definition;177
7.1.5.4;5.5.4 Summary of Algorithm;178
7.1.6;5.6 Modeling Chemical Reactions;178
7.1.6.1;5.6.1 Simple Chemical Reaction Model;179
7.1.6.2;5.6.2 Two-Species Lotka-Volterra Model;180
7.1.6.3;5.6.3 Bimolecular Reaction;181
7.1.7;5.7 Inferring Gene Regulatory Networks;182
7.1.7.1;5.7.1 The Small Artificial Gene Regulatory Network;184
7.1.7.2;5.7.2 The Large-Scale Artificial Gene Regulatory Network with;187
8;Part VI Conclusions and Future Research;189
8.1;6 Concluding Remarks and Further Research;190
8.1.1;6.1 Limitations of Conventional Computational Intelligence;190
8.1.2;6.2 Towards Tree-Structure Based Hierarchical Hybrid Computational Intelligence;191
8.1.2.1;6.2.1 Tree Structure Based Evolutionary Computation Models;191
8.1.2.2;6.2.2 Hierarchical Hybrid Computational Intelligence;191
8.1.3;6.3 Static and Dynamical Models;195
9;References;196