消费信息能用于个人信用风险评价吗  

Can consumer information be used in personal credit risk assessment

作　　者：温红梅[1] 隋昕[1] Wen Hongmei;Sui Xin(School of finance Harbin University of Commerce,Harbin 150000)

机构地区：[1]哈尔滨商业大学金融学院,哈尔滨150028

出　　处：《北方经贸》2021年第8期7-10,共4页Northern Economy and Trade

摘　　要：本文选取某生活服务类和现金贷为一体的平台借款人信息和还款表现数据,分别利用Logistic回归模型和Light GBM算法分析消费信息在个人信用风险评价体系中的识别能力。研究结果表明,消费场景数量、消费频次等级和消费能力等级三个消费信息变量对个人信用风险具有显著影响;加入消费信息能够有效提升模型准确性,提升个人信用风险的评价能力。在信用风险评价领域,消费信息能够补充传统征信信息无法覆盖的信息,提高个人信用风险+评价准确性。This paper selects the information and repayment performance data of a life service and cash loan platform,and analyzes the recognition ability of consumer information in the personal credit risk evaluation system by using Logistic regression model and LightGBM algorithm respectively.The results show that the number of consumption ability,consumptionfrequency,consumptionscenehave significant influence on personal credit risk.The accuracy of the model can be effectively improved by adding consumption information and the evaluation ability of individual credit risk.In the field of credit risk assessment,consumer information can supplement the information that traditional credit information cannot cover,and improve the accuracy of personal credit risk+evaluation.

关 键 词：消费信息 信用风险 信用风险评价 LOGISTIC LightGBM 

分 类 号：F830.5[经济管理—金融学]