Support Vector Regression for financial time series forecasting

Abstract

Recently, Support Vector Regression (SVR) has been a popular tool in financial time series forecasting. This study deals with the application of Support Vector Regression in stock composite index forecasting. A preprocessing method for accelerating support vector regression training is presented in this paper. Then we propose a method of support vector regression by modifying the regularized risk function. A data set from Shanghai Stock Exchange is used for the experiments to test the validity of our methods. © 2006 International Federation for Information Processing.