According to the idea of value investment,the classification model of stock investment value was built based on SVM(Support Vector Machines) in this paper.500 A-share stocks was randomly selected as sample firstly,and the financial data which distinct influent the stock investment value was selected to construct the sample feature set.Then the classification model of stock investment value was built based on SVM,the comparison of BP neural network and RBF neural network was proposed in the end.The results showed the classification effect and generalization capability was the best for the SVM.