摘要
稳定客户和吸引客户是移动通信企业提高竞争力的关键。基于大量实验数据将数据挖掘的决策树方法引入移动通信行业客户流失分析中,通过对数据的预处理,利用C4.5算法创建决策树,通过测试流失的与未流失的客户,平均正确识别率为91.6%。决策树体现的规则与经验基本一致,为移动通信企业建立客户流失的预警机制提供了决策支持。
Holding and attracting customers is the key to raise competitive of mobile communication company. Based on a lot of data, a kind of data mining algorithm C4.5 has been used in the analyzing of the customer churn. The paper makes the data pretreatment and makes decision tree by C4. 5 algorithm, from the testing results, the average rate of the correct recognition of the customers which have churned and not churned is 91.6 %. These rules from the decision tree are corresponded with the experiences, it can supply decision support for mobile communication company to establish the precaution and mechanism of customer churn.
出处
《计算技术与自动化》
2009年第3期98-101,共4页
Computing Technology and Automation
