【摘要】 以P2P网络借贷为例,从人人贷中选取2015—2018年共7 559条记录,通过数据挖掘模型来对借款人违约风险进行分析,并识别出影响借款人违约的主要因素,这些数据挖掘模型主要包括决策树、支持向量机和随机森林。主要结论包括:第一,运用数据挖掘模型来预测违约风险效果都很好,其中最好的是随机森林;第二,特征重要性程度前五依次为信用等级、借款金额、借款周期、借款利率、借款人所在企业的规模。更多还原

【Abstract】 Taking P2P network lending as an example,this paper selected 7559 records from personal loans from 2015 to 2018,analyzed borrowers default risk through data mining model,and identified the main factors affecting borrowers’ default.These data mining models mainly include decision tree,support vector machine and random forest.The main conclusions include:firstly,using data mining model to predict default risk is very good,the best of which is random forest;secondly,the top five characteristics of importance are credit rating,loan amount,loan cycle,loan interest rate and working time of borrowers.更多还原