摘要
为了有效地向手机用户提供换机服务,建立一种换机预测模型。利用孤立森林算法,排查与换机预测无关的异常电信用户。将排查后的数据集通过K-Medoids聚类分析精细划分为3个用户簇,利用SMOTE和Tomek组合采样的方法,处理每个用户簇的不平衡问题。最后将各个用户簇的数据通过XGBoost算法进行训练,并根据格式搜索法得出最优换机预测模型。实验结果表明,该换机预测模型的预测准确率高于其他预测模型,可较好地为电信用户提供换机服务。
A prediction model is proposed to effectively provide service to telecom users for changing mobile phones. In this model, the isolation forest algorithm is used to eliminate outliers in the user s datasets by comparing abnormal value from the algorithm with the threshold, and the K-mediods clustering method is used to get fine division of user clusters. The combination of synthetic minority oversampling technique (SMOTE) and Tomek sampling method is then used to deal with the unbalance problem in each user s cluster, and finally XGBoost algorithm is used in each user cluster datasets to train model for prediction and obtain the optimal model by grid search. Experimental results show that this prediction model has better prediction performance than other models and can provide better switch service for telecom users.
作者
卢光跃
吴洋
吕少卿
闫真光
LU Guangyue;WU Yang;LYU Shaoqing;YAN Zhenguang(Shaanxi Key Laboratory of Information Communication Network and Security,Xi'an University of Posts and Telecommunications, Xi'an 710121,China)
出处
《西安邮电大学学报》
2019年第2期94-97,104,共5页
Journal of Xi’an University of Posts and Telecommunications
基金
陕西省工业科技攻关计划资助项目(2015GY-013,2016GY-113)
陕西省教育厅科研计划资助项目(17JK0703)
