A Study Of Wavelet Support Vector Machine Theory And Its Applications In Forecasting Financial Index

Support vector machine is an efficient method for classification and prediction under small samples. It is a novel machine learning method with the rapid development in recent years. Especially, it is applied to financial time series prediction these days. Support vector machine deals with small samples, nonlinear, high dimension, local minima and other traditional problems well. However, the advantage of standard support vector machine is restricted in practice due to the ideal assumption of the sample distribution. In order to solve the practical problem according to the present situation, the novel support vector machine method in extreme conditions is researched by making the best of the distribution information of unknown status samples and the effective information in limited samples.Aiming at the low prediction accuracy due to the underutilization of the information in limited samples, we study multivariable prediction theory and wavelet support vector machine theory. According to [1], we know about wavelet support vector machine theory. Besides, it also discusses the application of wavelet support vector machine theory in Forecasting of Stock Returns. In this paper, we will try to find a way to combine multivariable prediction method and wavelet support vector machine theory. Then we use the new method to predict the financial index..