Research on Extreme Risk Warning for Financial Market Based on RU-SMOTE-SVM

Taking the Shanghai Stock Exchange Composite Index and Shenzhen Stock Exchange Component Index as the objects of research,this paper combines Random Under-Sampling(RU),Synthetic Minority Over-Sampling Technique(SMOTE)with Support Vector Machine(SVM) to establish an improvement SVM——RU-SMOTE-SVM,which is applied to predict the extreme risk in Chinese financial market and compared with conventional SVM,SMOTE-SVM,RU-SMOTE-NN and RU-SMOTE-DT.The result of investigation illustrates that RU-SMOTE-SVM not only outperforms conventional SVM,but also has a higher predictive accuracy than SMOTE-SVM,simultaneously,has a more excellent predictive performance than RU-SMOTE-NN and RU-SMOTE-DT.