基于信息增益与多元逐步回归组合技术的网贷平台借款人违约风险影响因素辨识

Identifying the Factors Affecting Credit Risk Borrowers of Online Loan Platform Based on Hybrid Technology of Information Gain and Multiple Stepwise Regression

本文以美国著名的Lending Club网贷平台为对象进行分析,通过构建信息增益与多元逐步回归组合技术探讨借款人违约风险的影响因素。研究结果显示,网贷平台借款人的借款金额、借款利率、还款额度、住房状况、逾期次数、公开信用账户、循环额度利用率、借款目的与借款人违约行为之间具有显著的影响。其中,借款金额、贷款利率、住房状况、公开信用账户对借款人违约风险存在显著的正向影响;借款人还款额度、借款目的、循环额度利用率、逾期次数对违约风险存在显著负向影响。然后,对网络借贷平台中各影响因素做了进一步分析,并根据研究结果提出了针对性建议,为网络借贷行业的风险防范提供了参考。

This paper analyzes the famous Lending Club online loan platform in the United States,and explores the influencing factors of borrower default risk by constructing information gain and multiple stepwise regression combination techniques.Results show that the borrowing amount,borrowing rate,repayment amount,housing status,number of overdue,public credit account,circulation quota utilization,borrowing purpose and borrower default behavior of the borrower of the online loan platform have a significant impact.Among them,the loan amount,loan interest rate,housing status,and public credit account have a significant negative impact on the default risk of the borrower;the borrower’s repayment amount,the purpose of borrowing,the utilization of the circulation quota,and the number of overdue have a significant positive impact on the default risk.Then,the reasons for the influencing factors in the network lending platform are further analyzed,and the targeted recommendations are put forward according to the research results,which provides a reference for the risk prevention of the network lending industry.