運用集成學習分類架構預測信用貸款購買行為

Predicting fiduciary loan purchasing behavior using ensemble learning model

10.6338/JDA.201412_9(6).0001

李天行(Tian-Shyug Lee)；陳怡妃(I-Fei Chen)；施讓龍(Ranglong Shi)；呂奇傑(Chi-Jie Lu)

集成學習 ； 分類技術 ； 信用貸款 ； 信用貸款購買行為 ； Ensemble learning ； classification ； fiduciary loan ； fiduciary loan purchasing behavior

Journal of Data Analysis

9卷6期（2014 / 12 / 01）

1 - 26

繁體中文

隨著銀行對消費金融業務的重視，信用貸款目前已是許多銀行在消費金融的重點項目之一，因此若能建立有效的信用貸款購買行為預測模式來協助銀行區分出有貸款意願與無貸款意願的消費者，有助於業者有效的找出信用貸款的目標顧客，降低行銷成本並提升銷售成功率。集成學習（ensemble learning）是一種藉由相互結合多種不同建模技術以處理一個特定問題的機器學習方法，其績效通常較單一學習器更佳與更穩健。本研究應用集成學習建構信用貸款消費購買行為預測模式，透過多數投票之方式整合羅吉斯迴歸（Logistic regression, LR）、區別分析（Discriminant analysis, DA），傳遞類神經網路（Back propagation neural network, BPN）、支援向量機（Support vector machines, SVM）及極限學習機（Extreme learning machine, ELM）等五種單一技術的分類結果，以獲得較穩健與準確的購買行為分類結果。本研究以某銀行之信用貸款資料為實證資料，在鑑別正確率、Kappa值及期望誤置成本等三個績效指標下，使用交叉驗證的方式來評估與比較所提之集成學習架構與其他五個模式的分類績效。實證結果顯示，在以每個方法於三種指標績效排名為基準的綜合比較下，集成學習之整體分類績效最佳，不僅有最高的鑑別正確率，也有最低的誤置成本。因此本研究所提之集成學習架構能有效的用於信用貸款購買行為預測，可協助企業未來在進行信用貸款行銷時辨別出具有購買意願客戶。

In this study, an ensemble learning model is proposed for predicting fiduciary loan purchasing behavior. Ensemble learning is the process by which multiple techniques/models are strategically combined to solve a particular issue. It is mainly applied to improve the performance of single model. In the proposed ensemble learning model, the classification results of the five well-known classification techniques including logistic regression (LR), discriminant analysis (DA), back propagation neural network(BPN), support vector machine(SVM) and extreme learning machine (ELM) are combined by majority voting to generate a better and robust classification model for fiduciary loan purchasing behavior prediction. The cross validation process and three performance indexes including correct classification rate, kappa value and misclassification cost are used to evaluate the performance of the proposed model. Experimental results from a real fiduciary loan data showed that the proposed ensemble learning model outperforms the five single classification methods in terms of correct classification rate and misclassification cost. It is an effective alternative for predicting fiduciary loan purchasing behavior.

1. Akgun, A.(2012).A comparison of landslide susceptibility maps produced by logistic regression, multi-criteria decision, and likelihood ratio methods: a case study at İzmir, Turkey.Landslides,9,93-106.
2. Dietterich, T. G.(2000).Ensemble methods in machine learning.Lecture Notes on Computer Sciences,1875,1-15.
3. Hansen, L. K.,Salamon, P.(1990).Neural network ensembles.IEEE Transactions on Pattern Analysis and Machine Intelligence,12(10),993-1001.
4. Hosmer, D. W., Jr,Lemeshow, S.(2004).Applied Logistic Regression.New York:John Wiley and Sons.
5. Huang, G. B.,Zhu, Q. Y.,Siew, C. K.(2006).Extreme learning machine: theory and applications.Neurocomputing,70,489-501.
6. Karaa, A.,krichene, A.(2012).Credit-risk assessment using support vectors machine and multilayer neural network models: A comparative study case of a Tunisian bank.Journal of Accounting and Management Information Systems,11,587-620.
7. Lai, K. K.,Yu, L.,Wang, S.,Wei, H.(2006).A novel nonlinear neural network ensemble model for financial time series forecasting.Lecture Notes in Computer Science,3991,790-793.
8. Liu, Y.,Yao, X.,Higuchi, T.(2000).Evolutionary ensembles with negative correlation learning.IEEE Transactions on Evolutionary Computation,4,380-387.
9. Lu, C. J.(2014).Sales forecasting of computer products based on variable selection scheme and support vector regression.Neurocomputing,128,491-499.
10. Lu, C.J.,Shao, Y. E.,Li, C.C.(2014).Recognition of concurrent control chart patterns by integrating ICA and SVM.Applied Mathematics and Information Sciences,8,681-689.
11. Mileris, R.(2010).Estimation of loan applicants default probability applying discriminant analysis and simple Bayesian classifier.Economics and Management,15,1078-1084.
12. Nagadevara, V.(2010).Hybrid models and error weighting for prediction customer churn in telecom industry.Review of Business Research,10,35-45.
13. Narzo, D.,Cocchi, D.(2010).A Bayesian hierarchical approach to ensemble weather forecasting.Journal of the Royal Statistical Society: Series C (Applied Statistics),59(3),405-422.
14. Polikar, R.(2006).Ensemble based systems in decision making.Circuits and Systems Magazine, IEEE,6,21- 45.
15. Rodríguez, J. J.,Kuncheva, L. I.,Alonso, C. J.(2006).Rotation forest: A new classifier ensemble method.IEEE Transactions on Pattern Analysis and Machine Intelligence,28,1619-1630.
16. Traiteur, J. J.,Callicutt, D. J.,Smith, M.,Roy, S. B.(2012).A short-term ensemble wind speed forecasting system for wind power applications.Journal of Applied Meteorology and Climatology,51,1763-1774.
17. Vapnik, V.(2000).The Nature of Statistical Learning Theory.New York:Springer.
18. Wang, G.,Hao, J.,Ma, J.,Jiang, H.(2011).A comparative assessment of ensemble learning for credit scoring.Expert systems with applications,38,223-230.
19. Zhang, G.,Patuwo, B. E.,Hu, M. Y.(1998).Forecasting with artificial neural networks: The state of the Art.International Journal of Forecasting,14,35-62.
20. 林宜錩(2014)。碩士論文(碩士論文)。朝陽科技大學財務金融系。
21. 林岳伸(2013)。碩士論文(碩士論文)。雲林科技大學財務金融系。
22. 林盟凱(2004)。碩士論文(碩士論文)。世新大學資訊管理學系。
23. 林蔚蒲(2013)。碩士論文(碩士論文)。國立中山大學機械與機電工程學系研究所。
24. 莊士玄(2009)。碩士論文(碩士論文)。國立中央大學產業經濟研究所。
25. 莊子妙(2013)。碩士論文(碩士論文)。國立高雄應用科技大學金融資訊研究所。
26. 陳怡安(2010)。碩士論文(碩士論文)。銘傳大學國際企業學系。
27. 彭慧雯(2001)。碩士論文(碩士論文)。國立台北科技大學生產系統工程與管理研究所。
28. 楊佳穎(2005)。碩士論文(碩士論文)。崑山科技大學企業管理研究所。
29. 葉建良(2006)。碩士論文(碩士論文)。輔仁大學應用統計學研究所。
30. 趙李英記(1999)。碩士論文(碩士論文)。中國文化大學資訊管理學系。
31. 蔡佩芬(2012)。碩士論文(碩士論文)。中臺科技大學健康產業管理研究所。
32. 謝弘一(2011)。博士論文(博士論文)。輔仁大學商學研究所。
33. 藍于絢(2009)。碩士論文(碩士論文)。國防大學管理學院資訊管理學系。
34. 蘇容嫻(2014)。碩士論文(碩士論文)。國立臺北大學統計學系。
35. 温堯欽(2013)。碩士論文(碩士論文)。國立聯合大學資訊管理學系。