應用資料探勘於信用卡授信決策模式之實證研究

Constructing a Credit Diagnosis Model Using Data Mining Technology--A Study of Credit Card Industries

林芝儀(Chin-Yi Lin)

指導教授 ： 鄭春生

元智大學/工程學院/工業工程與管理學系/碩士(2003年)

摘要

近年來，台灣金融機構爭相投入信用卡的戰場中，為提高發卡量、爭取發卡時效及減少審核時間，導致信用卡逾期比率提高、催收成本與呆帳件數大幅成長。因此，建立信用風險管理與良好的授信政策是目前發卡銀行首要重視的問題。
本研究以國內某金融機構為研究對象，以資料探勘的方式從資料中找出有用的資訊。研究中先以交叉分析與相關分析了解資料結構，進而運用資料探勘技術中的倒傳遞類神經網路與決策樹分別建構信用卡決策模式，有效預測申請者之客戶類型。研究發現倒傳遞類神經網路的預測能力達97.2%，決策樹之預測能力達93.3%。本研究根據兩種資料探勘技術的特色與預測能力，結合實務上授信專家的經驗，建構出一套信用卡授信決策模式，以人機界面的方式產生即時預警效果，並判斷是否核卡及授權適當額度以減少呆帳，更可針對不同類型的客戶採取合適的服務。藉由此系統，發卡機構可隨時餵養異常呆帳資料，讓決策系統藉由不斷的學習，並修正學習機制，達到更即時，更精確之效果。

關鍵字

信用風險管理 ； 資料探勘 ； 倒傳遞類神經網路 ； 決策樹

並列摘要

In recent years, the financial institutions in Taiwan try to join the battlefield of credit card. To increasing the amount of cardholders and decreasing the examining
time leads to the higher overdue rate of credit card and the more receivables on demand. Therefore, establishing the correct credit risk management and fitted model of credit scoring is the most important issue for these issuer banks.
Data mining is the art of finding patterns in data and is a new approach based on a general recognition that there is undraped value in large databases and utilities data-driven extraction of information. This research uses stratified random sampling method to choose the training and testing samples. Firstly, we use cross analysis and correlation analysis to discover the data structure. Secondly, we use back-propagation neural network and decision tree to obtain the knowledge of different consumer types. In this study, the forecasting accuracy of back-propagation neural network is 97.2% and decision tree is 93.3%. It is concluded that using back-propagation neural network in building a customer classification model is more appropriate than decision tree. The model may assist banks’ administrators using their applicants’ demographics to distinguish their risk attitude for approving an appropriate credit limit for a cardholder’s expenditure to promote the total credit card profit for banks.

並列關鍵字

credit risk management ； data mining ； back-propagation neural network ； decision tree

參考文獻

21.郭敏華，債信評等，智勝，2000。

5.Han, J. and M. Kamber, Data Mining: Concepts and Techniques, Morgan Kaufmann, CA, USA (2001).

8.Johnson, R.A. and D.W. Wichern, , Applied Multivariate Statistical Analysis, Prentice Hall (2002).

9.Quinlan, J. R., “Induction of decision trees,” Mach. Learn., 1(1), 81-106 (1986).

10.Quinlan, J. R., C4.5: Programs for Machine Learning, Morgan Kaufmann, San Mateo, CA, USA (1993).