By Topic

Credit Risk Assessment Model of Commercial Banks Based on Support Vector Machines

Sign In

Cookies must be enabled to login.After enabling cookies , please use refresh or reload or ctrl+f5 on the browser for the login options.

Formats Non-Member Member
$31 $13
Learn how you can qualify for the best price for this item!
Become an IEEE Member or Subscribe to
IEEE Xplore for exclusive pricing!
close button

puzzle piece

IEEE membership options for an individual and IEEE Xplore subscriptions for an organization offer the most affordable access to essential journal articles, conference papers, standards, eBooks, and eLearning courses.

Learn more about:

IEEE membership

IEEE Xplore subscriptions

3 Author(s)
Xin-Ying Zhang ; Dept. of Manage. Sci. & Eng., Harbin Inst. of Technol., Harbin, China ; Chong Wu ; Ferretti, A.-F.

Scope: Commercial banks, as the key of the nation's economy and the center of financial credit, play a multiple irreplaceable role in the financial system. Credit risks threaten the economic system as a whole. Therefore, predicting bank financial credit risks is crucial to prevent and lessen the incoming negative effects on the economic system. Objective: This study aims to apply a credit risk assessment model based on support vector machines (SVMs) in a Chinese case, after analyzing the credit risk rules and building a credit risk system. After the modeling, it presents a comprehensive computational comparison of the classification performances of the techniques tested, including back-propagation neural network (BPN) and SVMs. Method: In this empirical study, we utilize statistical product and service solutions (SPSS) for the factor analysis on the financial data from the 157 companies and Matlab and Libsvm toolbox for the experimental analysis. Conclusion: We compare the assessment results of SVMs and BPN and get the indication that SVMs are very suitable for the credit risk assessment of commercial banks. Empirical results show that SVMs are effective and more advantageous than BPN. SVMs, with the features of simple classification hyperplane, good generalization ability, accurate goodness of fit, and strong robustness, have a better developing prospect although there are still some problems with them, such as the space mapping of the kernels, the optimizing scale, and so on. They are worthy of our further exploration and research.

Published in:

Management and Service Science, 2009. MASS '09. International Conference on

Date of Conference:

20-22 Sept. 2009