基于支持向量机的商业银行信用风险度量模型  

Credit Risk Measure Model of Chinese Commercial Bank Based on SVM

作　　者：余晨曦[1] 梁潇[2] 

机构地区：[1]武汉理工大学管理学院,武汉430070 [2]武汉理工大学计算机科学与技术学院,武汉430070

出　　处：《计算机与数字工程》2008年第11期10-14,共5页Computer & Digital Engineering

摘　　要：运用商业银行信贷风险的度量及管理理论,提出了对商业银行信贷信用风险管理中的贷款违约判别方法,并运用统计学习理论在模式识别领域的研究成果—支持向量机技术,构建了基于支持向量机的我国商业银行信贷信用风险度量模型。将支持向量机的非线性分类器应用到贷款违约的判别中,应用上市公司的财务数据进行计算,并且将其结果与多元线性判别分析的结果进行对比。得出了支持向量机在对贷款违约的判别中有很好的判别效果的结论。This thesis utilizes the commercial bank credit risk measurement and the management theory, proposed to the commercial bank credit risks management of loan defaults discriminant method. And using the research results of statistical learning theory in the field of pattern recognitio-SVM （Support vector machines） technology, constructed the SVM - Based Chinese commercial bank credit ＆ credit loan risk measure model. And the SVM nonlinear classification is applied to loan defaults discriminant, application of the financial data of listed companies to calculate, and compared its result with the multiple linear distinction analysis＇s result. Obtained the conclusion that support vector machines to have the very good distinction effect to the loan defaults discrimination.

关 键 词：商业银行 信贷风险 信用风险 支持向量机 

分 类 号：TP393.09[自动化与计算机技术—计算机应用技术]