摘要
本文采用大数据相关思维为指导,采用相关数据挖掘技术,对商业银行电子商务平台零售客户的风险计量模型进行设计和构建。首先根据客户相关信息及行为变量,采用决策树方法将电子商务客户分为若干群体,并分析总结了不同客户群体的行为特征;然后通过分析电子商务平台客户的相关行为信息以及其在线下的金融产品交易活动,构建相关建模备选变量;在此基础上,采用国际先进银行通用的建模方法,以零售B2C类的消费客户为例,进行了风险评分模型实证分析。结果表明,通过大数据分析基础上的电子商务客户风险模型构建,具有较好的风险识别能力和区分度,各项检验结果较为合理。本研究对商业银行利用大数据思维构建风险管理工具以及提高电子商务平台的风险管理水平提供了实证支持,具有一定现实参考意义。
In this paper, risk measurement model of retail customers, who use e-commerce platform of commercial bank, is designed and built by using ideas on big data and techniques on data mining. First, according to the customer-related information and behavioral variables, e-commerce customers are classified into several groups based on the decision tree method. The behavior characters of different customer groups are also analyzed and summarized. Then, alternative variables of modeling are generated by examining the behavior of e-commerce platform users and their trading activities on financial products. Employing the modeling techniques widely adopted by international leading banks, the empirical analysis on risk scoring model is conducted by taking B2C consumer customers as an example. The results show that the risk model design of e-commerce customers based on big data analysis is capable of both identifying and distinguishing the risk with reasonable test results. This study which provides empirical evidence for the commercial banks to build risk management tools and improve risk management of e-commerce platform has realistic significance.
出处
《投资研究》
北大核心
2014年第4期16-26,共11页
Review of Investment Studies
