Application Of Data Mining Method In The Evaluation Index System Of Financial Risk In Listed Corporation

Facing the increasingly brutal competition of the market environment,how to evaluate the enterprise financial risk objectively and warn it timely,so as to improve the risk management level of enterprises effectively is gradually becoming an important problem for managers to pay more attention to.With the rapid development of data mining technology,and the application of relevant theories are becoming more and more mature,through establishing evaluation index system,using data mining method to mine the company's financial data deeply,so as to study the condition of financial risk has become more and more scholars'research subject.Data mining method has lots of advantages that traditionl statistical analysis methods don't have.Firstly,we don't have to set up a lot of assumptions,just adjust the threshold,and it can explore the rules of the data automatically.Moreover,data mining technology needn't too complex mathematical theory of reasoning,just through the data filter,it can find the hidden features of the database.Therefore,data mining method is applied in this paper to analyze the the financial risk of the enterprise,to establish dynamic enterprise financial risk evaluation index system model.Using data mining methods such as association rules to find the high trust degree rules,using artificial neural network to warn the crisis of the enterprise financial risk timely.The main innovative contributions of this dissertation are as follows:1.By establishing the analysis flow chart of enterprise financial risk,the new process of financial risk analysis are proposed for enterprise in the new information environment.2.The pre-processing method are proposed by combining clustering analysis and statistical analysis,comprehensive application of difference test,multiple linear test,correlation analysis and clustering analysis techniques.Through the access of artificial neural network model,the results show that the data pre-processing method proposed in this paper improves the accuracy of the model's judgement significantly.