Nonlinear system identification based on LSSVM within the evidence framework

Support vector machine (SVM) is a new learning machine based on the statistical learning theory. A regression algorithm based on least squares support vector machine (LS SVM) within the Bayesian evidence framework is discussed. Also the Gauss kernel parameter selecting method is proposed. Under the evidence framework, the regularization and kernel parameters can be adjusted automatically, which can achieve a fine tradeoff between the minimum error and modelpsilas complexities. This method is applied to nonlinear system identification and the simulation results show the effectiveness and superiority of the proposed approach. It provides a new way for modeling and identification of complicated industrial processes.