A rapid parameter selection method of kernel function

Parameter selection is an important issue for kernel method. As a commonly used parametric search method, grid search can ensure to obtain good parameters, but requires large amount of computation. Aiming to the problem, a rapid hybrid parameter selection method of kernel function based on uniform design and grid search is proposed in this paper. Uniform design is used to preliminary search of a well-performed parameter combination in the value range. Grid search is used to search around the parameter combination to improve the performance of kernel function further. Effectiveness and generality of the method are proved by experiment.