Research On The Application Of Data Mining In Listed Company's Credit Risk In China

With the rapid growth of credit in today's global economy, problems related to credit riskhave attracted much attention. Credit risk has become one of the important risks whichfinancial institutions have to face. With the rapid growth of credit in today's global economy,problems related to credit risk have attracted muchattention.Credit risk has become one of theimportant risks which financial institutions have to face. How to get an accurate estimation ofthe credit risk plays a key role in the decision making of financial intermediaries,investorsand government pay close attention to.Chinese security market is a big and complex market. It is significant not only forChinese Securities Regulatory Commission, Shanghai Stock Exchange and Shenzhen StockExchange as supervisors, but also for listed companies themselves, obliges and mass investorsto construct a scientific system to evaluate listed companies'credit condition. At the sametime, the Data Mining techniques and methods continue to mature and provide a just resolvethe severe credit risk management problems.Purpose of this paper is to make the great advantage of data mining techniques applied toresearch into credit risk, credit risk of listed companies, data processing, methods of analysisfor new thinking and exploration, data processing using data mining to make up for theadvantages of credit risk data deficiencies. By analyzing the characteristics of credit risk data,makes data mining method for listed companies in the application of credit risk provide strongsupport.The paper analyzes the causes of listed companies'credit risk, takes Chinese mainlandlisted companies as research objects, and regards special treatment as a signal of credit crisis;at last it selected 135 annual financial statements of 135 listed companies in 2005 as thesample studying. Using the Spss Clementine data mining tool selected variables by principal component analysis and Logistic regression model analysis and forecast. This article alsoutilization in the financial static data, dynamic stock market data, combined with targetsrelated to non-financial factors, to our listed companies to measure the credit risk of empiricalresearch. Positive researches indicated that the prediction correct rate of sample is higher, soresearch approach is effective and reasonable.