The Study On Kernels In Support Vector Machine

Support Vector Machine(SVM) is the main content of Statistics Learning Theory developed from 1990s. The kernel function is the crucial ingredient of SVM. In these kernel functions, many researchers attach importance to Gauss kernel function because of its peculiar property and broad applications.First, this paper introduces VC dimension theory and structural risk minimization principle. Uses Gauss kernel function through SVM and discusses its separability and the property of localization. Then we explain the importance of selecting parameter in SVM and select Gauss kernel radius and punishment-parameter C.Next, through local kernel function and global kernel function, we mix two kinds of kernels together on the base of its main property.Last, we propose a corrected Gauss kennel function, and apply in Voice Active Detection experiments, computered simulation results show us that the new kennel function has a better performance than Gauss kennel function.The related software is developed on Matlab6.5 and VC++6.0.