Customer churn prediction in telecom based on random forest algorithm

Abstract

In recent years, with the in-depth implementation of the telcom's Number Portability work, there are more and more Number Portability users, and the Number Portability users have become a major factor of customer churn. In order to reduce customer churn, the machine learning algorithm of Python analyzes the characteristics of outbound users, establishes customer churn model prediction, locates transfer subscribers in advance, and maintains stability retention, effectively reducing the probability of outbound users.