Research on the performance of SVM with Fourier-Kernel function and application on regression

Contrast to traditional methods, Support Vector Machine (SVM) has better performance on generalization. It has widely applications on pattern recognition.. but less on regression now. And the common choice of kernel function is Radial Basis Function, so few Studies on other special kernels. In this paper, the performance of SVM based on Fourier kernel is studied which aims at the regression in signal processing problems, and the influence of parameter q on performance of SVM is analyzed. A conclusion is drawn that the integral of Fourier kernel in one period is a constant and the concept of equivalent kernel function width is proposed. At last, Simulation verifies that SVM based on Fourier kernel has better performance than the one,based on RBF kernel in the field of signal processing.