摘要
为了更好地控制借款人的信用风险,利用支持向量机对个人信用进行预测与分析,在支持向量机对个人信用评估产生缺陷的基础上提出基于代价敏感学的CART决策树预测个人信用的方法。实证分析表明:该方法能够较好地对借款人信用状况进行预测,为互联网金融机构进行相关风险管理提供理论依据。
To predict and analysis individual credit by using support vector machine ( SVM), the author puts forward a method of personal credit evaluation approach based on cost-sensitive CART, which provides a theoretical basis to commercial banks of the assessment for personal credit status about related risk management.
出处
《成都工业学院学报》
2016年第4期60-62,71,共4页
Journal of Chengdu Technological University
关键词
支持向量机
个人信用
互联网金融机构
CART决策树
风险管理
Support Vector Machine (SVM)
personal credit
online financing
classification and regression tree
risk management