贝叶斯决策树在客户流失预测中的应用

Bayesian decision tree applying in forecasting customer churn

针对电信企业客户流失问题,提出采用贝叶斯决策树算法的预测模型,将贝叶斯分类的先验信息方法与决策树分类的信息熵增益方法相结合,应用到电信行业客户流失分析中,分别将移动公司的客户数据以及UCI数据纳入到模型中得出相应的结果。加入贝叶斯节点弥补决策树不能处理缺失值以及二义性数据的缺点。检验结果表明,基于贝叶斯推理的决策树算法在牺牲了较小的训练时间与分类时间的情况下,得到了比仅基于决策树算法更高的覆盖率与命中率。

According to telecom enterprise customer churn problem, a prediction model based on Bayesian decision tree algorithm is put forward. It combines the prior information method of Bayesian classification and the information gain method of decision tree classification, and applies to the analysis of telecom enterprise customer churn, then puts the customer data of mobile company and data UCI in the model respectively and gets the relevant results. Added Bayesian node to make up for the decision tree cannot handle the missing value and the ambiguity data. The applications indicate that, Bayesian decision tree algorithm sacrifices a little training time and classification time, but gets higher coverage rate and hit rate than basic decision tree method.