生存分析在电信客户流失中的应用  

Application of survival analysis in telecom customer churn

作　　者：张三妞 张智斌[1] ZHANG Sanniu;ZHANG Zhibin(Kunming University of Science and Technology,Kunming 650504,China)

机构地区：[1]昆明理工大学,云南昆明650504

出　　处：《现代电子技术》2023年第12期84-90,共7页Modern Electronics Technique

摘　　要：近年来电信行业发展迅猛,电信市场趋于饱和,在目前的市场环境下,预测电信客户的流失及流失时间有利于挽留客户,减少客户损失。针对电信客户流失的预测,文中采用KM与Cox的电信客户生存分析模型预测电信客户流失的时间,将KM分析模型用于电信客户流失的组间比较;运用Cox风险比例模型对多因素进行分析,全面预测电信客户的生存时间。通过构建生存分析模型可实现对潜在流失客户及客户流失时间的预测,从而制定相应的有效方案挽留目标。In recent years,the telecom industry has developed rapidly,and the telecom market has become saturated.In the current market environment,it is beneficial for retaining customers and reducing customer losses by predicting the loss of telecommunications customers and the time of loss.For the prediction of telecom customer churn,the survival analysis models of KM and Cox are used to predict the time of telecom customer churn.The KM analysis model is used for comparison among groups of telecom customer churn,and the Cox risk ratio model is used to perform multi⁃factor analysis,so as to comprehensively predict the survival time of telecom customer.By constructing a survival analysis model,it is possible to predict potential lost customers and their loss time,thereby enacting corresponding effective schemes to retain the customers.

关 键 词：电信客户流失 生存分析 KM分析模型 Cox风险比例模型 多因素分析 生存时间预测 

分 类 号：TN98-34[电子电信—信息与通信工程]