Application of SVM in Financial Research

Based on the structural risk minimization, support vector machine is a new method of data mining. Since it has effectively solved complicated problems of classification and prediction, it has been widely used in many cross-disciplinary fields. This paper has reviewed and analyzed SVMpsilas application to the classification and prediction in the financial field. It has a promising future of applying to company´s credit rating, early warning, stock prices forecast and so on. However, we hold that the correct selection of kernel and different sub-assembly function, as well as parameters, is the key point to optimize the application of SVM.