改进的RFM模型和K-means算法在会员分类中的应用研究  

Research on the Application of Improved RFM Model and K-Means Algorithm in Membership Classification

作　　者：张利斌[1] ZHANG Libin(School of Intelligent Equipment,Changzhou College of Information Technology,Changzhou 213164,China)

机构地区：[1]常州信息职业技术学院智能装备学院,江苏常州213164

出　　处：《常州信息职业技术学院学报》2024年第3期47-51,共5页Journal of Changzhou College of Information Technology

摘　　要：针对传统RFM模型用于会员分类会产生失真的问题,对RFM模型提出了改进,增加了客户关系长度和客户购买周期两个参数。同时针对传统的K-means算法存在的问题,提出了一种基于样本对象特征方差加权与中心初始化的K-means算法。利用改进的RFM模型对会员进行分类,可以有效地提高分类效率。Aiming at the distortion caused by traditional RFM models used for member classification,this paper proposes an improvement to the RFM model by adding two parameters:customer relationship length and customer purchase cycle.At the same time,a K-means algorithm based on sample object feature weighting and central initialization is proposed to address the problems of traditional K-means algorithms.Using the improved RFM model for member classification can effectively improve classification efficiency.

关 键 词：RFM模型 K-MEANS聚类 会员分类 

分 类 号：TP311.13[自动化与计算机技术—计算机软件与理论]