Search

In order to improve the accuracy and efficiency of customer churn prediction of railway scattered freight, according to the loss characteristics of railway scattered freight customers, this paper proposed a customer churn identification method based on CDL model. On this basis, facing the problem of big data, it proposed a C4. 5 decision tree customer churn prediction model based on Hadoop parallel framework. Simulation results show that the model has good accuracy and predictive ability, and as the number of samples increases, obviously improves the efficiency of Hadoop parallel framework, and doesn’t affect the accuracy and prediction ability of churn prediction model. [ABSTRACT FROM AUTHOR]