逻辑回归和支持向量机在客户信用分类中的应用  

Application of Logistic Regression and Support Vector Machine in Customer Credit Classification

作　　者：代雯月 DAI Wen-yue(School of Mathematics,Sichuan University of Arts and Science,Dazhou 635000,China)

机构地区：[1]四川文理学院数学学院,达州635000

出　　处：《价值工程》2023年第5期139-141,共3页Value Engineering

基　　金：分类模型在信用评分领域的应用研究,四川文理学院科研启动基金资助。

摘　　要：本文借助R语言,基于逻辑回归和支持向量机模型理论,分别建立客户存款信用分类模型来判断个人客户是否订阅定期存款的状况,通过对比两种模型的错判率和评估性能,选取出相对最优的逐步逻辑回归模型,刻画出最潜在的客户画像,预测订阅存款的可能性,进而对测试集数据分类和判别。这将作为商业银行界用来判断客户是否订阅定期存款的依据,同时也利于银行工作人员选择出更好的营销对策。With the help of R language,based on the theory of logical regression and support vector machine model,this paper establishes customer deposit credit classification models to determine whether individual customers subscribe to fixed deposits.By comparing the error rate and evaluation performance of the two models,the paper selects the relatively optimal stepwise logical regression model,depicts the most potential customer profile,predicts the possibility of subscription deposits,and then classifies and distinguishes the test set data.This will serve as the basis for the commercial banking industry to judge whether customers subscribe to fixed deposits,and also help the bank staff to choose better marketing strategies.

关 键 词：分类模型 逻辑回归 支持向量机 客户分类 性能评估 

分 类 号：TP181[自动化与计算机技术—控制理论与控制工程]