Adaptive consumer credit classification

• Published in: The Journal of the Operational Research Society Vol. 63; no. 12; pp. 1645 - 1654
• Main Authors: Pavlidis, N G, Tasoulis, D K, Adams, N M, Hand, D J
• Format: Journal Article
• Language: English
• Published: London Taylor & Francis 01-12-2012; Palgrave Macmillan; Palgrave Macmillan UK; Taylor & Francis Ltd
• Subjects: Applied sciences; Business and Management; Classification; Consumer credit; Credit; Credit risk; Credit scoring; Datasets; Decision theory. Utility theory; Default; Discriminant analysis; Exact sciences and technology; General Paper; General Papers; H-measure; Logistic regression; Management; Mathematical vectors; online learning; Operational research and scientific management; Operational research. Management science; Operations research; Operations Research/Decision Theory; Population; Population distributions; population drift; Population dynamics; Portfolio theory; Regression analysis; Studies; Test bias; measure; population drift; logistic regression; credit scoring; online learning; Loans; Finance; Financial market; Risk analysis; Economic sciences; Adaptive method; Standards; Selection problem; Logistic regression; Credit; Scoring credit; Classification; H-measure; Potential method; Economic crisis
• ISSN: 0160-5682; 1476-9360
• DOI: 10.1057/jors.2012.15

Credit scoring methods for predicting creditworthiness have proven very effective in consumer finance. In light of the present financial crisis, such methods will become even more important. One of the outstanding issues in credit risk classification is population drift. This term refers to changes occurring in the population due to unexpected changes in economic conditions and other factors. In this paper, we propose a novel methodology for the classification of credit applications that has the potential to adapt to population drift as it occurs. This provides the opportunity to update the credit risk classifier as new labelled data arrives. Assorted experimental results suggest that the proposed method has the potential to yield significant performance improvement over standard approaches, without sacrificing the classifier's descriptive capabilities.