基于可变距与可变频策略的终端换机模型

Terminal Replacement Model Based on Variable Distance and Variable Frequency Box Preprocessing Strategy

预测潜在换机用户是实现终端精准营销的关键。为了挖掘通讯公司大数据仓库中的数据价值,完善"大数据+渠道"的先进管理营销体系,文章的模型构建基于数据挖掘技术,在通讯公司积累的大数据中分析出与终端换机行为相关度高的特征数据,包括用户基础信息、用户长期消费习惯信息、用户对应终端配置信息数据等多个维度数据,综合业务场景与数据分布特点对数据进行可变距和可变频分箱预处理操作,并利用融合多周期后验的分类算法构建模型。通过模型评估,结果表明本模型可大幅度提高预测精准度。

Prediction of potential users is the key to achieve terminal precision marketing.In order to mine the data value in the big data warehouse of the communication company,and improve the advanced management and marketing system of"big data+channels".In this paper,the model built based on data mining technology,in the communications company accumulated large data analysis and terminal replacement behavior characteristic of high correlation data,including user basic information,the user long-term consumption habits,corresponding terminal user configuration information data multiple dimensional data,such as integrated business scenarios and data distribution characteristic of data for variable distance and variable frequency points data preprocessing,and use the classification algorithm of posterior fusion more cycle model of the building.Through model evaluation,the results show that this model can greatly improve the precision of prediction.