A total least squares proximal support vector classifier for credit risk evaluation
In this paper, a total least squares (TLS) version of proximal support vector machines (PSVM) is proposed for credit risk evaluation. The formulation of this new model is different from the original PSVM model, so a novel iterative algorithm is proposed to solve this model. A simulation test is first implemented on a classic two-spiral dataset, and then an empirical experiment is conducted on two publicly available credit datasets. The experimental results show that the proposed total least squares PSVM (TLS-PSVM) is at least comparable with PSVM and better than other models including standard SVM model.
KeywordsTotal least squares method Proximal support vector machine Credit risk evaluation
This work is partially supported by grants from the National Science Fund for Distinguished Young Scholars (NSFC No. 71025005) and the National Natural Science Foundation of China (NSFC No. 90924024).
- Frank A, Asuncion A (2010) UCI machine learning repository, University of California Irvine. http://archive.ics.uci.edu/ml
- Fu G, and Hu G (2008) Total least square support vector machine for regression. In: Proceedings of the 2008 international conference on intelligent computation technology and automation, vol 1, pp 271–275Google Scholar
- Fung G, and Mangasarian OL (2001) Proximal support vector machine classifiers. In: Proceedings of international conference of knowledge discovery and data mining, pp 77–86Google Scholar
- Hardle W, Moro RA, Schafer D (2005) Predicting bankruptcy with support vector machines. In: Statistical tools for finance and insurance. Springer, Berlin, pp 225–248Google Scholar
- Mathworks (2008), MATLAB R2008aGoogle Scholar
- Rakotomamonjy A (2004) Optimizing area under ROC curve with SVMs, ECAI 04 ROC and Artificial Intelligence Workshop, 20–27 August, Valencia, SpainGoogle Scholar
- Suykens JAK, van Gestel T, De Brabanter J, De Moor B, and Vandewalle JV (2002) Least squares support vector machines. World Scientific, SingaporeGoogle Scholar
- Vapnik V (1995) The nature of statistical learning theory. Springer, New YorkGoogle Scholar
- Vapnik V (1998) Statistical learning theory. Wiley, New YorkGoogle Scholar
- Van Gestel T, Baesens B, Garcia J, van Dijcke P (2003) A support vector machine approach to credit scoring. Bank en Financiewezen 2:73–82Google Scholar