The Applications Of Data-mining In The Evaluation Of Individual Credit Risk

In recent years, with the rapid development of Chinese economy, the incomelevel of urban and rural residents continue to improve, the personal investment andconsumption of residents also continue to increase, the personal credit business offinancial institutions has a rapid development, the individual credit risk is becomingthe focus concerned by the banks and other financial institutions; Meanwhile, with thecontinuous improvement of personal credit information system, the personal creditbehavior is being recognized as a necessary means to promote the level of personalmorality, and to maintain social and economic order. Thus, the evaluation ofindividual credit risk is of great significance, no matter for commercial banks or theresidents.However, the reality is that when the system of evaluation of individual creditrisk has entered the stage of specialization, even the stage of industrialization in thewestern countries, the construction of our country’s system of evaluation of individualcredit risk has just started. These problems, for example, the index of evaluationsystem is imperfect, the applicability of the model is not strong, lack of effectiveevaluation system in banking practices and so on, have been plagued by the domesticfinancial theorists and practitioners.This paper firstly describes the concept of individual credit risk and Data-Mingtechnology in details, and clarify the reasons why this paper choice the Logisticregression. Then based on comprehensiveness, importance, scientific, impartiality andoperational these principles, selects the alternative indicators from six aspects related tothe evaluation of individual credit risk, including the basic information, careerinformation, solvency and stability of the situation and so on; selects the personalcustomers’ data of N commercial bank, selects the indexes from the alternative indicatorsby the information gain value; transfers the categorical variables into continuousvariables by using of the WOE, in order to build a evaluation model of individual creditrisk by Logistic regression. Finally, when the evaluation model of individual creditrisk has been built, this paper constructs the N Commercial Bank’s individual creditrisk evaluation system by the J2EE technology.