In the space, the idea of seeking s

In the space, the idea of seeking solutions in the case of linear separability is obtained by its optimal classification surface, which is reflected by the decision function of the original space classifier. However, the feature dimension disaster of the mapping may occur. In order to solve this problem, the concept of SVM kernel function is introduced, and the function calculation in the inner product original space of the transformed transformation space is adopted, which guarantees the good classification performance of the nonlinear problem of SVM kernel function, and also solves the dimension disaster problem of the nonlinear mapping.