A new fuzzy support vector machine to evaluate credit risk
Research output: Journal Publications and Reviews (RGC: 21, 22, 62) › 21_Publication in refereed journal › peer-review
Author(s)
Related Research Unit(s)
Detail(s)
Original language | English |
---|---|
Pages (from-to) | 820-831 |
Journal / Publication | IEEE Transactions on Fuzzy Systems |
Volume | 13 |
Issue number | 6 |
Publication status | Published - Dec 2005 |
Link(s)
Abstract
Due to recent financial crises and regulatory concerns, financial intermediaries' credit risk assessment is an area of renewed interest in both the academic world and the business community. In this paper, we propose a new fuzzy support vector machine to discriminate good creditors from bad ones. Because in credit scoring areas we usually cannot label one customer as absolutely good who is sure to repay in time, or absolutely bad who will default certainly, our new fuzzy support vector machine treats every sample as both positive and negative classes, but with different memberships. By this way we expect the new fuzzy support vector machine to have more generalization ability, while preserving the merit of insensitive to outliers, as the fuzzy support vector machine (SVM) proposed in previous papers. We reformulate this kind of two-group classification problem into a quadratic programming problem. Empirical tests on three public datasets show that it can have better discriminatory power than the standard support vector machine and the fuzzy support vector machine if appropriate kernel and membership generation method are chosen. © 2005 IEEE.
Research Area(s)
- Classification, Credit scoring, Fuzzy theory, Machine learning, Support vector machine (SVM)
Citation Format(s)
A new fuzzy support vector machine to evaluate credit risk. / Wang, Yongqiao; Wang, Shouyang; Lai, K. K.
In: IEEE Transactions on Fuzzy Systems, Vol. 13, No. 6, 12.2005, p. 820-831.Research output: Journal Publications and Reviews (RGC: 21, 22, 62) › 21_Publication in refereed journal › peer-review