Contents -- Preface -- Evolutionary Computation and its Applications -- 1. Maximal Margin Algorithms for Pose Estimation Ying Guo and Jiaming Li -- Contents -- 1.1. Introduction -- 1.2. Pose Detection Algorithm -- 1.2.1. Procedure of pose detection -- 1.2.2. Eigen Pose Space -- 1.3. Maximal Margin Algorithms for Classification -- 1.3.1. Support vector machines -- 1.3.2. Boosting -- 1.3.3. Soft margin AdaBoost -- 1.3.4. Margin distribution -- 1.4. Experiments and Discussions -- 1.4.1. Data preparation -- 1.4.2. Data preprocessing -- 1.4.3. Experiment results of Group A -- 1.4.4. Margin distribution graphs -- 1.4.5. Experiment results of Group B -- 1.5. Conclusion -- References -- 2. Polynomial Modeling in a Dynamic Environment based on a Particle Swarm Optimization Kit Yan Chan and Tharam S. Dillon -- Contents -- 2.1. Introduction -- 2.2. PSO for Polynomial Modeling -- 2.3. PSO vs. GP -- References -- 3. Restoration of Half-toned Color-quantized Images Using Particle Swarm Optimization with Multi-wavelet Mutation Frank H.F. Leung, Benny C.W. Yeung and Y.H. Chan -- Contents -- 3.1. Introduction -- 3.2. Color Quantization With Half-toning -- 3.3. Formulation of Restoration Algorithm -- 3.3.1. PSO with multi-wavelet mutation (MWPSO) -- 3.3.2. The tness function -- 3.3.3. Restoration with MWPSO -- 3.3.4. Experimental setup -- 3.4. Result and Analysis -- 3.5. Conclusion -- References -- Fuzzy Logics and their Applications -- 4. Hypoglycemia Detection for Insulin-dependent Diabetes Mellitus: Evolved Fuzzy Inference System Approach S.H. Ling, P.P. San and H.T. Nguyen -- Contents -- 4.1. Introduction -- 4.2. Hypoglycemia Detection System: Evolved Fuzzy Inference System Approach -- 4.2.1. Fuzzy inference system -- 4.2.1.1. Fuzzification -- 4.2.1.2. Fuzzy reasoning -- 4.2.1.3. Defuzzification -- 4.2.2. Particle swarm optimization with wavelet mutation.

4.2.2.1. Wavelet mutation -- 4.2.3. Choosing the HPSOWM parameters -- 4.2.4. Fitness function and training -- 4.3. Results and Discussion -- 4.4. Conclusion -- References -- Neural Networks and their Applications -- 5. Study of Limit Cycle Behavior of Weights of Perceptron C.Y.F. Ho and B.W.K. Ling -- Contents -- 5.1. Introduction -- 5.2. Notations -- 5.3. Global Boundness Property -- 5.4. Limit Cycle Behavior -- 5.5. Application of Perceptron Exhibiting Limit Cycle Behavior -- 5.6. Conclusion -- References -- 6. Artificial Neural Network Modeling with Application to Nonlinear Dynamics Yi Zhao -- Contents -- 6.1. Introduction -- 6.2. Model Structure -- 6.3. Avoid Overfitting by Model Selection -- 6.3.1. How it works -- 6.3.2. Case study -- A. Computational experiments -- B. Experimental data -- 6.4. Surrogate Data Method for Model Residual -- 6.4.1. Linear surrogate data -- 6.4.2. Systematic flowchart -- 6.4.3. Identification of model residual -- 6.4.4. Further investigation -- 6.5. The Diploid Model Based on Neural Networks -- 6.6. Conclusion -- Acknowledgments -- References -- 7. Solving Eigen-problems of Matrices by Neural Networks Yiguang Liu, Zhisheng You, Bingbing Liu and Jiliu Zhou -- Contents -- 7.1. A Simple Recurrent Neural Network for Computing the Largest and Smallest Eigenvalues and Corresponding Eigenvectors of a Real Symmetric Matrix -- 7.1.1. Preliminaries -- 7.1.2. Analytic solution of RNN -- 7.1.3. Convergence analysis of RNN -- 7.1.4. Steps to compute 1 and n -- 7.1.5. Simulation -- 7.1.6. Section summary -- 7.2. A Recurrent Neural Network for Computing the Largest Modulus Eigenvalues and Their Corresponding Eigenvectors of an Anti-symmetric Matrix -- 7.2.1. Preliminaries -- 7.2.2. Analytic solution of RNN -- 7.2.3. Convergence analysis of RNN -- 7.2.4. Simulation -- 7.2.5. Section summary.

7.3. A Concise Recurrent Neural Network Computing the Largest Modulus Eigenvalues and Their Corresponding Eigenvectors of a Real Skew Matrix -- 7.3.1. Preliminaries -- 7.3.2. Analytic solution -- 7.3.3. Convergence analysis of RNN -- 7.3.4. Simulation -- 7.3.5. Comparison with other methods and discussions -- 7.3.6. Section summary -- 7.4. A Recurrent Neural Network Computing the Largest Imaginary or Real Part of Eigenvalues of a General Real Matrix -- 7.4.1. Analytic expression of z (t) 2 -- 7.4.2. Convergence analysis -- 7.4.3. Simulations and discussions -- 7.4.4. Section summary -- 7.5. Conclusions -- References -- 8. Automated Screw Insertion Monitoring Using Neural Networks: A Computational Intelligence Approach to Assembly in Manufacturing Bruno Lara, Lakmal D. Seneviratne and Kaspar Althoefer -- Contents -- 8.1. Introduction -- 8.2. Background -- 8.2.1. The screw insertion process: modelling and monitoring -- 8.2.2. Screw insertion process: monitoring -- 8.3. Methodology -- 8.3.1. Screw insertion signature classification: successful insertion and type of failure -- (a) Single insertion case -- (b) Multiple insertion cases -- (c) Multiple output classifications -- 8.3.2. Radial basis function neural network for error classification -- 8.3.3. Simulations and experimental study -- 8.4. Results of Simulation Study -- 8.4.1. Single insertion case -- 8.4.2. Generalization ability -- 8.4.3. Multi-case classification -- 8.5. Results of Experimental Study -- 8.5.1. Single insertion case -- 8.5.2. Generalization ability -- 8.5.3. Four-output classification -- 8.6. Conclusions -- Acknowledgments -- References -- Support Vector Machines and their Applications -- 9. On the Applications of Heart Disease Risk Classification and Hand-written Character Recognition using Support Vector Machines S.R. Alty, H.K. Lam and J. Prada -- Contents -- 9.1. Introduction.

9.1.1. Introduction to support vector machines -- 9.1.2. The maximum-margin classifier -- 9.1.3. The soft-margin classifier -- 9.1.4. Support vector regression -- 9.1.5. Kernel functions -- 9.2. Application: Biomedical Pattern Classification -- 9.2.1. Pulse wave velocity -- 9.2.2. Digital volume pulse analysis -- 9.2.3. Study population and feature extraction -- 9.2.3.1. Physiological features -- 9.2.3.2. Signal-based features -- 9.2.4. Results -- 9.3. Application: Hand-written Graffiti Recognition -- 9.3.1. Data acquisition and feature extraction -- 9.3.2. SVR-based recognizer -- 9.3.3. Simulation results -- 9.4. Conclusion -- Acknowledgements -- References -- Appendix A -- Appendix B -- Appendix C -- Appendix D -- Appendix E -- 10. Nonlinear Modeling Using Support Vector Machine for Heart Rate Response to Exercise Weidong Chen, Steven W. Su, Yi Zhang, Ying Guo, Nghir Nguyen, Branko G. Celler and Hung T. Nguyen -- Contents -- 10.1. Introduction -- 10.2. SVM Regression -- 10.3. Experiment -- 10.4. Data Analysis and Discussions -- 10.5. Conclusion -- References -- 11. Machine Learning-based Nonlinear Model Predictive Control for Heart Rate Response to Exercise Yi Zhang, Steven W. Su, Branko G. Celler and Hung T. Nguyen -- Contents -- 11.1. Introduction -- 11.2. Background -- 11.2.1. Model-based predictive control (MPC) -- 11.2.1.1. MPC structure -- 11.2.2. Dynamic matrix control (DMC) -- 11.3. Control Methodologies Design -- 11.3.1. Discrete time model -- 11.3.2. Switching control method -- 11.3.3. Demonstration of tuned DMC parameters for control system of cardio-respiratory response to exercise -- 11.3.4. Simulation -- 11.4. Conclusions and Outlook -- References -- 12. Intelligent Fault Detection and Isolation of HVAC System Based on Online Support Vector Machine Davood Dehestani, Ying Guo, Sai Ho Ling, Steven W. Su and Hung T. Nguyen -- Contents.

12.1. Introduction -- 12.2. General Introduction on HVAC System -- 12.3. HVAC Parameter Setting -- 12.4. HVAC Model Simulation -- 12.5. Fault Introduction -- 12.6. Parameter Sensitivity -- 12.7. Incremental Decremental Algorithm of SVM -- 12.8. Algorithm of FDI by Online SVM -- 12.9. FDI Simulation -- 12.10. Conclusion -- References -- Index.