为了解决移动客户流失量建模与预测中的一些难题,结合移动客户流失量的变化特点,提出一种基于数据挖掘的移动客户流失量预测算法.首先收集移动客户流失量的历史样本,并通过预处理消除一些无用样本,然后根据贝叶斯决策树算法对移动客户类型进行分类,最后针对具体的移动客户预测流失量.结果表明,该算法建模速度优于其他移动客户流失量预测模型,可以获得更优的移动客户流失量预测结果.
In order to solve problems of mobile customer churn modeling and prediction,and combined with change characteristics of mobile customers churn,this paper put forward a mobile customer loss prediction model based on Bayesian decision tree algorithm.Firstly,mobile customer loss amount of history data are collected,and pretreated to eliminate some useless samples,and then mobile customer types are classified according to Bayesian decision tree algorithm,finally,the model is applied into specific mobile customer churn prediction.The results show that modeling speed of the proposed is superior to other mobile customer churn prediction models,and can get better prediction results of mobile customer churn.