基于BP神经网络的有线电视用户流失预测

With the rapid development of science and technology, the traditional cable TV business suffered the impact of telecom operators and Internet free video. In this case, cable TV users gradually began to churn, cable TV operators should timely identify and retain the churn of users. In this paper, a BP neural network based cable TV user churn prediction method is proposed. First of all, this paper analyzes and processes the basic information of real users and viewing information data. Then according to the characteristics of cable TV users, select the parameters that can reflect the characteristics of churn and conduct dimensionality reduction processing on them. Finally, a churn prediction model is established based on BP neural network. The result of the example model proves that the cable TV user churn prediction method based on BP neural network has satisfactory prediction effect.