Mining Potential 5G Mobile Users Based on Logistic Regression and Random Forest

All operators in the world are actively deploying 5G networks and promoting 5G communication consumption packages in many countries. In recent years, in the field of communication, data mining technology has provided strong support for accurate marketing. Based on Logistic Regression and Random Forest model, this paper analyzes the data of potential users of 5G mobile. Firstly, according to the consumption characteristics of 5G users, we process the 5G user data set in the dimensions of user identification, consumption behavior information and over-set information. Secondly, we built Logistic Regression and Random Forest to mine 5G potential users. Finally, we compare the experimental results of the above two models according to the ROC curve. The experimental results show that the Random Forest model can accurately locate the potential users of 5G and bring economic benefits to operators for rapid promotion and accurate mar-keting.