A Research on Constructing a Credit Card Risk Management Alert Model

• Main Authors: Lion-hua Liu, 劉玲華
• Other Authors: Deng-Yuan Huang
• Format: Others
• Language: zh-TW
• Published: 2009
• Online Access: http://ndltd.ncl.edu.tw/handle/03647630500195160536

碩士 === 輔仁大學 === 應用統計學研究所 === 97 === Credit card is one of the most important economic commodities for a bank. Therefore, how to lower and control the risk on credit card transactions become particularly important. Also, to prevent credit card theft is the prime principle and direction to lower transaction loss. The purpose of this study is to establish a credit card risk management model. We use the decision tree (CART) method to observe legal and illegal credit card transaction, then use Logistic regression to establish a credit card risk model. As new transactions occurred, we can use this model to filter out high risk consumption pattern and this can save more time on the whole procedure. Result shows, consumption of pattern, country, and on-site transaction are with statistical significance, whereas transaction amount is not.