支持向量机由Vapnik首先提出，像多层感知器网络和径向基函数网络一样，支持向量机可用于模式分类和非线性回归，该程序主要实现svm的分类和回归功能。(SVM was first proposed by Vapnik. Like multilayer sensor network and radial basis function network, SVM can be used for pattern classification and non-linear regression. The program mainly implements svm classification and regression functions.)