Hilbert空间映射下非线性分类的变元可分离核函数

SEPARABLE-VARIABLES KERNEL FUNCTION USED TO NONLINEAR CLASSIFICATION UNDER HILBERT SPACE MAPPING

分析变元可分离函数K(u,v)=k(u)k(v)成为核函数对非线性支持向量分类机的作用,由新方法构造的非线性支持向量分类机可以使相应的算法得以简化.

It is concluded that separable variables function K=（u,v）=k（u）k（v） can be used as kernel function to provide a new method in selecting new kernel function of nonlinear classification machines based on support vector machine.Using nonlinear classification support vector machines structured by the new method,the relevant algorithm will be simplified.