Research On Financial Risk Early Warning Of Listed Companies Based On SVM Model

Starting from the major financial risks faced by listed companies,this paper selects the companies facing delisting risks due to financial distress in China's stock market in 2016,and selects the proportion of 1:2 as the object of study,and selects the solvency,profitability,profitability,operating ability and growth ability of the enterprises.The financial indicators,such as the 2014 annual report as a training set,the 2015 sample as a test set,based on factor analysis method to build a SVM model for financial risk early warning model,the classification accuracy rate is 82%,early warning accuracy rate is 77.25%.Secondly,the introduction of TOPSIS method after the dual classification model optimization test.The results show that the SVM model constructed by TOPSIS method can significantly improve the accuracy of the model,the classification accuracy increased to 95.2381%,the early warning accuracy increased to 82.6596%,which shows that the double classification can effectively improve the early warning accuracy.