Based on kernel principal component analysis combined kernel function algorithm

This paper presents a novel combination of kernel function of support vector machine in construction.This combination of kernel functions,Gauss kernel function and polynomial kernel function respective features together,build a both interpolation and extrapolation properties of kernel function.After the experimental verification,the kernel function is applied in the kernel principal component analysis method,which can effectively improve the recognition accuracy and efficiency.