Font Size: a A A

Research On The Applications Of Data Mining In Credit-Risk Management Of Credit Card

Posted on:2009-04-20Degree:MasterType:Thesis
Country:ChinaCandidate:Y ChenFull Text:PDF
GTID:2189360272490013Subject:Quantitative Economics
Abstract/Summary:PDF Full Text Request
Development of the credit card industry has brought the huge opportunity for our country's banks. At the same time, the commercial banks are facing unprecedented challenges. Whether or not it has a strong ability of balancing risks and benefits is the key to success for our banks. It is to be urgent to enhance commercial bank's credit card risk management ability. Advanced data mining techniques can mine some unknown and valuable rule from the mass data of credit card. It is no doubt to provide a strong technical support for credit card credit-risk management.This paper researches on credit-risk measurement of credit card. It introduces data mining methods which are the most advanced techniques in current data analysis into the credit-risk management of credit card.At the beginning, the paper discusses the concepts and characteristics of credit card, credit card risk and credit card risk management, and their related theories. It elaborates the theory of data mining, the basic principle of rough set specially and their applications in credit-risk management of credit card. And it compares three methods of credit card credit-risk measurement, including rough set method, discriminant analysis method and Logistic regression model by the theoretical and empirical researches. The paper shows that rough set is valid in credit-risk management of credit card, and gives some useful suggestions.The innovative points of this paper are:Firstly, this paper introduces rough set method into credit card credit-risk management of our country. It discusses the basic principle of rough set and its applicability in credit card credit-risk management.Secondly, it does empirical analysis of rough set in credit card credit-risk management using Taiwan data, and compares the empirical results of rough set, discriminant analysis and Logistic regression model using the same data. As a result, the effect of rough set method is best of all. Thirdly, it shows that rough set method is valid in credit card credit-risk management by the theoretical and empirical results. The paper provides some useful suggestions and a new technique of credit card credit-risk management for our country.
Keywords/Search Tags:Credit Card, Credit Risk, Data Mining, Rough Set
PDF Full Text Request
Related items