Prediction For Stock Market Based On Support Vector Machine

Stock market is a complex non-linear dynamic system. It is very difficult to develop the inherent rules using the traditional timing prediction technology. This paper use the support vector machine to predict the stock market, based on much analysis on the traditional methods and on methods upon the artificial neural networks. This paper puts forward a parameter selection strategy for stock market predicton. And introduce the steps of prediction of the stock market. Meanwhile, this paper puts forward a parameter selection method for nonlinear Support Vector Machine, and study the effect of the parameter of the kernel function on the prediction. This paper applies the standard Support Vector Machine to two SSE indices and some stock to make prediction; the result show the support vector machine is effective method. Meanwhile, this paper compares the support vector machine with the BP neural networks, the result shows the BP neural networks is better than support vector machine on the training dataset but on test dataset, the support vector machine is better.