A Hybrid Support Vector Regression for Time Series Prediction

Abstract

Due to time series forecasting involves a rather complex data pattern, there are lots of novel forecasting approaches to improve the forecasting accuracy. Unlike most conventional neural network models, which are based on the empirical risk minimization principle, SVM applies the structural risk minimization principle to minimize an upper bound of the generalization error, rather than minimizing the training error. It has been successfully used to solve nonlinear regression and times series problems. One particular model can not capture all data patterns easily. This article presents a hybrid SVM model with mixed kernel function to exploit the unique strength of the linear and nonlinear SVM models to deal with this problem. Furthermore, parameters of both the linear and nonlinear SVM models are determined by Immune Algorithm (IA). A numerical example is employed to compare the performance of the proposed model. Experiment results reveal that the proposed model promising alternative for forecasting time series problems.