"信息增益与多元逐步回归技术对网贷平台借款人违约风险的影响分析及启示.pdf"

With the rise of the Internet, online lending platforms have become a popular alternative to traditional financial institutions. In this study, the factors influencing borrower default risk on online lending platforms were analyzed using a combination of information gain and multiple stepwise regression techniques. The empirical analysis was conducted using data from Lending Club, a well-known online lending platform. The results of the study revealed that factors such as loan grade, loan amount, interest rate, employment length, income verification, annual income, debt-to-income ratio, total loan revolving balance, and length of credit history significantly impact borrower default risk on online lending platforms. On the other hand, factors such as loan term, employment length, income verification, annual income, debt-to-income ratio, total loan revolving balance, and length of credit history were found to have no significant impact on borrower default risk. Based on the research findings, recommendations for institutional design and legal regulation of online lending platforms in China were proposed. These recommendations aim to promote the healthy and sustainable development of the online lending market in China, which is of great theoretical and practical significance. Keywords: Information Gain, Multiple Stepwise Regression, Default Risk, Lending Club, Wanfang Data