lssvm工具箱1_8.pdf下载

Matlab R2009b - R2013a: LS-SVMlab1.8 - Linux and Windows (32 and 64 bit): lssvm工具箱使用说明1——8英文原版 Support Vector Machines (SVM) is a powerful methodology for solving problems in nonlinear classification, function estimation and density estimation which has also led to many other recent developments in kernel based learning methods in general [14, 5, 27, 28, 48, 47]. SVMs have been introduced within the context of statistical learning theory and structural risk minimization. In the methods one solves convex optimization problems, typically quadratic programs. Least Squares Support Vector Machines (LS-SVM) are reformulations to standard SVMs [32, 43] which lead to solving linear KKT systems. LS-SVMs are closely related to regularization networks [10] and Gaussian processes [51] but additionally emphasize and exploit primal-dual interpretations. Links between kernel versions of classical pattern recognition algorithms such as kernel Fisher discriminant analysis and extensions to unsupervised learning, recurrent networks and control [33] are available. Robustness, sparseness and weightings [7, 34] can be imposed to LS-SVMs where needed and a Bayesian framework with three levels of inference has been developed [44]. LS-SVM alike primal-dual formulations are given to kernel PCA [37, 1], kernel CCA and kernel PLS [38]. For very large scale problems and on-line learning a method of Fixed Size LS-SVM is proposed [8], based on the Nystr¨om approximation [12, 49] with active selection of support vectors and estimation in the primal space. The methods with primal-dual representations have also been developed for kernel spectral clustering [2], data visualization [39], dimensionality reduction and survival analysis [40]