基于SVM的房贷信用评估的应用研究  

Application research of housing loan credit evaluation based on SVM

作　　者：王波[1] 郝艳友[2] 刘勇奎[1] 刘爽[1] 

机构地区：[1]大连民族学院计算机科学与工程学院,辽宁大连116605 [2]大连理工大学计算机科学与工程系,辽宁大连116001

出　　处：《计算机工程与设计》2008年第19期5110-5113,共4页Computer Engineering and Design

摘　　要：信贷风险是金融机构风险主要来源。支持向量机(SVM)在解决两类问题上是一种较好的分类方法,其学习模型有较强的稳定性。对SVM在房贷信用评估应用中的问题进行了研究和解决,如核函数选取,参数选取,样本非均衡问题等。实验得出在实际应用中径向基模型较好,采用Grid-search方法调整参数,能达到更好的推广能力和预测结果,用分别惩罚支持向量机能有效解决样本非均衡问题。试验结果也证明了基于SVM的房贷信用评估方法优于原有的打分方法。Credit risk is the primary source of risk to financial institutions. Support vector machine （SVM） is a good classifier to solve binary classification problem and the learning results possess stronger robustness. Resolving the problem of the application of SVM in housing loan credit evaluation, such as the choice of kernel function and parameters, the problem of unbalance data. The experiment show RBF model is suitable better for practical application, grid-search method adjusts these penalty parameters to achieve better gene- ralization performances in the application, respectively penalty SVM （RP_SVM） can resolve unbalance problem efficiency. And it is proved that credit evaluation method based SVM is better than the old method.

关 键 词：信用评估 支持向量机 网格搜索 参数选取 非均衡样本 

分 类 号：TP181[自动化与计算机技术—控制理论与控制工程]