基于分类技术的信用评分模型研究

Study on the Credit Score Model Based on Classification Technology

本文基于数据挖掘和信用评分理论,对比了常见的4种分类模型技术,结合实证背景,从可解释性的角度说明了逻辑回归模型可行性。选取德国信用数据集,在逐步逻辑回归的基础上建立了信用评分卡模型,量化各用户的信用得分。

Based on data mining and credit score theory,this paper compares four common classification model techniques,combined with the empirical background,and explains the feasibility of logistic regression model from the perspective of interpretability.The German credit data set is selected,and the credit score card model is established on the basis of gradual logical regression to quantify the credit score of each user.