基于随机森林的潜在5G用户预测分析

Analysis of Potential 5G User Prediction Based on Random Forest Z

阐述基于收集的12 000条用户消费数据信息,包括4 000条5G用户和8 000条4G用户,利用随机森林算法提升潜在5G用户的识别率算法,从而预测用户未来是否有5G订阅需求。实验结果表明,基于随机森林的5G潜在用户的预测准确率、精确率和Recall分别达到了0.88、0.89和0.88。

This paper describes the algorithm of using random forest algorithm to improve the recognition rate of potential 5G users based on collected 12000 user consumption data information,including 40005G users and 80004G users,in order to predict whether users will have 5G subscription needs in the future.The experimental results show that the prediction accuracy,accuracy,and Recall of 5G potential users based on random forests have reached 0.88,0.89,and 0.88,respectively.