Contents -- Preface -- Chapter 1 Introduction -- 1.1 Multilayer perceptron neural networks -- 1.2 Regression and classification -- 1.3 Learning and generalization -- 1.3.1 Weight decay and effective number of parameters -- 1.3.2 Ridge regression -- 1.3.3 Bayesian learning -- 1.4 Principles of pattern recognition -- 1.4.1 Bayes rule and optimal classifier under Gaussian assumptions -- 1.4.2 Receiver operating characteristic -- 1.5 Dimensionality reduction methods -- 1.6 Parametric versus non-parametric approaches and RBF networks -- 1.7 Feedforward versus recurrent network models -- Chapter 2 Support Vector Machines -- 2.1 Maximal margin classification and linear SVMs -- 2.1.1 Margin -- 2.1.2 Linear SVM classifier: separable case -- 2.1.3 Linear SVM classifier: non-separable case -- 2.2 Kernel trick and Mercer condition -- 2.3 Nonlinear SVM classifiers -- 2.4 VC theory and structural risk minimization -- 2.4.1 Empirical risk versus generalization error -- 2.4.2 Structural risk minimization -- 2.5 SVMs for function estimation -- 2.5.1 SVM for linear function estimation -- 2.5.2 SVM for nonlinear function estimation -- 2.5.3 VC bound on generalization error -- 2.6 Modifications and extensions -- 2.6.1 Kernels -- 2.6.2 Extension to other convex cost functions -- 2.6.3 Algorithms -- 2.6.4 Parametric versus non-parametric approaches -- Chapter 3 Basic Methods of Least Squares Support Vector Machines -- 3.1 Least Squares Support Vector Machines for classification -- 3.2 Multi-class formulations -- 3.3 Link with Fisher discriminant analysis in feature space -- 3.3.1 Linear Fisher discriminant analysis -- 3.3.2 Fisher discriminant analysis in feature space -- 3.4 Solving the LS-SVM KKT system -- 3.4.1 Conjugate gradient iterative methods -- 3.4.2 LS-SVM classifiers: UCI benchmarking results.

3.5 Least Squares Support Vector Machines for function estimation -- 3.6 Links with regularization networks and Gaussian processes -- 3.6.1 RKHS and reproducing property -- 3.6.2 Representer theorem and regularization networks -- 3.6.3 Gaussian processes -- 3.7 Sparseness by pruning -- Chapter 4 Bayesian Inference for LS-SVM Models -- 4.1 Bayesian inference for LS-SVM classifiers -- 4.1.1 Definition of network model parameters and hyperparameters -- 4.1.2 Bayes rule and levels of inference -- 4.1.3 Probabilistic interpretation of LS-SVM classifiers (Level 1) -- 4.1.4 Inference of the hyperparameters (Level 2) -- 4.1.5 Inference of kernel parameters and model comparison (Level 3) -- 4.1.6 Design of the LS-SVM classifier in the Bayesian evidence framework -- 4.1.7 Example and benchmarking results -- 4.2 Bayesian inference for LS-SVM regression -- 4.2.1 Probabilistic interpretation of LS-SVM regressors (Level 1): predictive mean and error bars -- 4.2.2 Inference of hyperparameters (Level 2) -- 4.2.3 Inference of kernel parameters and model comparison (Level 3) -- 4.3 Input selection by automatic relevance determination -- Chapter 5 Robustness -- 5.1 Noise model assumptions and robust statistics -- 5.2 Weighted LS-SVMs -- 5.3 Robust cross-validation -- 5.3.1 M-Estimators -- 5.3.2 L-Estimators -- 5.3.3 Efficiency-robustness trade-off -- 5.3.4 A robust and efficient cross-validation score function -- Chapter 6 Large Scale Problems -- 6.1 Low rank approximation methods -- 6.1.1 Nystrom method -- 6.1.2 Incomplete Cholesky factorization -- 6.2 Fixed Size LS-SVMs -- 6.2.1 Estimation in primal weight space -- 6.2.2 Active selection of support vectors -- 6.3 Basis construction in the feature space -- 6.4 Combining submodels -- 6.4.1 Committee network approach -- 6.4.2 Multilayer networks of LS-SVMs.

Chapter 7 LS-SVM for Unsupervised Learning -- 7.1 Support Vector Machines and linear PCA analysis -- 7.1.1 Classical principal component analysis formulation -- 7.1.2 Support vector machine formulation to linear PCA -- 7.1.3 Including a bias term -- 7.1.4 The reconstruction problem -- 7.2 An LS-SVM approach to kernel PCA -- 7.3 Links with density estimation -- 7.4 Kernel CCA -- 7.4.1 Classical canonical correlation analysis formulation -- 7.4.2 Support vector machine formulation to linear CCA -- 7.4.3 Extension to kernel CCA -- Chapter 8 LS-SVM for Recurrent Networks and Control -- 8.1 Recurrent Least Squares Support Vector Machines -- 8.1.1 From feedforward to recurrent LS-SVM formulations -- 8.1.2 A simplification to the problem -- 8.1.3 Example: trajectory learning of chaotic systems -- 8.2 LS-SVMs and optimal control -- 8.2.1 The N-stage optimal control problem -- 8.2.2 LS-SVMs as controllers within the N-stage optimal control problem -- 8.2.3 Alternative formulation and stability issues -- 8.2.4 Illustrative examples -- Appendix A -- Bibliography -- List of Symbols -- Acronyms -- Index.