A new SVM parameter selection method based on the bound of structure risk

A new method of SVM parameter selection based on the bound of structure risk is proposed. First, the bound of the structure risk is theoretically analyzed. Then, the simulated experiment with the UCI data is designed. Comparisons are made between the proposed method and public databases UCI SVM parameter selection simulation. The results show that the method of boundary structural risk classification based on the second-class SVM model and multi-class classification SVM model is an effective, and less time consuming.