潜在流失客户的发掘模型探讨  

The mining model of defecting customers

作　　者：马辉民[1] 尹汉斌[1] 郭潇[2] 

机构地区：[1]华中科技大学管理学院 [2]西华师范大学数学系

出　　处：《华中科技大学学报（自然科学版）》2003年第9期28-30,共3页Journal of Huazhong University of Science and Technology(Natural Science Edition)

基　　金：国家自然科学基金资助项目 (70 0 710 11) .

摘　　要：提出了一个基于数据挖掘技术的解决方案 ,首先运用C4 .5决策树算法识别客户的客观指标集合 ,然后对这些客观指标进行偏差分析 。This paper provided a solution to the customer defection based on data mining techniques. C4.5 decision tree algorithm was used to identify objective indicators of customer, then the deviation of these indicators was detected. Finally, a time series model was constructed to predict the defecting customers. This article applied data mining correlative techniques in analyzing potential cosing customers, and could help enterprises to make decision of customer retention.

关 键 词：客户保持 偏差分析 数据掺掘 客户流失 

分 类 号：F274[经济管理—企业管理] TP311[经济管理—国民经济]