数据融合建模在信用卡申请评分模型中的应用

On the Application of Data Fusion Modeling in the Scoring Model of Credit Card Application

从背景、建模思路、模型指标体系构建,以及实际建模过程中的样本选择、建模数据安全防范、数据处理和变量特征分析、模型训练及检验、模型比较和模型结果应用等方面,描述如何使用信用卡申请表数据、支付类数据、互联网信贷类数据,进行数据融合建模,来解决传统信用卡申请评分模型辨识精度低及排序性差的问题,以覆盖更多的互联网申请用户和减低不良率。

This article describes how to conduct data fusion modelling by making use of credit card application form data,payment-type data,and internet credit data from the aspects of background,modelling ideas,and construction of model index system,sample selection in the actual modelling process,modelling data security,data processing and variables feature analysis,model training and testing,model comparison,and the application of model results,etc.It intends to solve the problem of low identification accuracy and poor sorting of the traditional credit card application scoring model,cover more internet application users,and reduce the non-performing rate.