论文部分内容阅读
预测移动终端用户是否有换机意向并向用户推荐和销售移动终端,已成为电信运营商的一项重要新业务,它不仅为公司带来了巨大利润,还提高了用户的满意度和忠诚度.移动用户换机预测通过学习用户在使用手机过程中产生的历史数据,来预测今后用户可能的换机趋向.在结合专家经验并统计大量用户数据的基础上,利用数据挖掘手段对用户数据进行特征分析,选取有效属性并对其进行区间划分,最后使用基于优势关系粗糙集方法对用户进行换机意向分类选择,以提高换机预测准确性.通过对用户数据集进行测试实验,证明了该方法对于预测移动用户换机的有效性.
Predicting whether mobile end-users have the intent to reboot and recommend and sell mobile terminals to subscribers has become an important new service for telecom operators, which not only brings huge profits to the company, but also improves user satisfaction and loyalty Mobile subscriber forecasting by predicting the possible trend of renewal of subscribers in the future by learning the historical data generated by the users during using mobile phones.Based on the experience of experts and the statistics of a large amount of user data and the data mining method, Feature analysis, select valid attributes and partition them, and finally use the rough set based on dominant relationship to classify the users in order to improve the predictive accuracy of the switch-in. By testing the user data set, Method for predicting the validity of mobile user replacement.