基于粗糙熵的电信客户行为推理研究

摘要

In telecommunications churn management , it is important and challenging to predict customers'churn by using telecommunication customer behavior data .With behavioral attributes and state property of telecom customers being delineated , telecom customers'behavior characteristics are analyzed .Then, based on rough entropy , an algorithm is presented to analyze the telecom customers'behavior .Meanwhile , an empirical analysis based on the behavioral data of a Hubei telecom enterprise is carried out .At last, asso-ciation rules between the behavioral attributes and state property of telecom customers are given .It shows that, by the proposed algorithm , telecommunication customers'behavior data can be effectively utilized to predict the customers'status.%如何利用电信客户的行为数据来对电信客户的流失进行预警一直是电信客户流失管理的重点与难点。对电信客户的行为特点进行了分析，给出了电信客户的行为属性以及状态属性，提出了基于粗糙熵的电信客户行为推理算法，并对湖北某电信运营商的客户行为数据进行了实证分析，得到了电信客户行为属性与状态之间的关联规则。证明了该算法可以有效地利用电信客户的行为数据对客户的状态进行有效的知识推理。