摘要
分析了电信行业客户关系管理系统的数据独有特点,提出基于客户细分的客户流失预测模型。首先,采用模糊核C-均值聚类算法用于客户细分并对细分结果进行分析,发现高价值客户的群体特征。再利用企业历史数据建立基于SAS数据挖掘技术的客户流失预测模型。最后,把高价值客户作为预测目标数据应用于该模型当中预测出有流失倾向的客户。实验结果表明,该方法有效可行,可以为企业提供准确、有流失倾向的客户名单。
The unique features of the customer relationship management（CRM） system in telecom industry is explored and a customer-churn model based on customer segmentation is proposed.Firstly,fuzzy kernel c-means clustering algorithm is used to segment customer and conclude high value customer group characteristics.Then,using the history data builds a prediction model of customer-churn based on SAS data mining technology.The result of customer segmentation is applied to customer-churn model and is gotten accuracy list of lost customer.Experiment proves this method obtain a satisfactory result of customer-churn.
出处
《计算机工程与设计》
CSCD
北大核心
2009年第24期5755-5758,共4页
Computer Engineering and Design
