摘要
准确的客户流失预测能够使移动业务针对特定用户需求和消费水平进行差异化服务。本文提出了一种基于贝叶斯网络的客户流失预测模型,根据案例分析的结果识别客户离网倾向原因,寻找具有高离网倾向的客户群,通过分类的特征分析,得出客户流失规则,并验证模型的有效性。
Accurate prediction of customers loss could enable mobile business to offer different service for specific needs and consumption of users. A model for the forecast of the loss of customers had been constructed based on Bayesian network. And according to the results of case studies, the paper identified the reasons for customer chum tend to look for a customer base with a high tendency to off-grid, and through the characteristics analysis of the classification, the rules of loss of customers had been arrived at and the effectiveness of the model had been verified.
出处
《电信科学》
北大核心
2009年第3期70-75,共6页
Telecommunications Science
基金
国家自然科学基金资助项目(No.70631003和No.90718037)
关键词
客户流失
贝叶斯网络
移动业务
预测
customer loss, Bayesian network, mobile service, prediction
