优化的K-means聚类算法在客户细分中的应用研究  

Research and application of customer segmentation model based on optimized K-means algorithm

作　　者：唐欣 TANG Xin(School of Mathematics and Information Science,North Minzu University,Yinchuan 750021,China)

机构地区：[1]北方民族大学数学与信息科学学院,银川750021

出　　处：《智能计算机与应用》2023年第9期194-196,F0003,共4页Intelligent Computer and Applications

摘　　要：传统的K-means聚类算法虽然操作简单快捷,但因随机选取聚类中心等问题容易陷入局部最优,导致算法不稳定。本文从样本间的关系出发,利用样本密度来优化K-means算法,并利用聚类有效性指标进行比较,优化后的K-means算法更具有稳定性且聚类准确率更高。最后,将该算法应用到客户细分RFM模型中,依据聚类结果找到适合不同消费者的营销策略,从而帮助企业更好地为其提供差异化、个性化服务。Although the traditional K-means clustering algorithm is simple and fast,it is easy to fall into local optimization due to the random selection of clustering centers,which leads to instability of the algorithm.This paper proposed an optimized K-means algorithm with the relationship of datasets and compared it with traditional K-means by clustering effectiveness index.It is concluded that the optimized K-means algorithm is more stable and has higher clustering accuracy.Finally,this algorithm is employed to the customer segmentation RFM model and give different market strategies with different customers by the clustering results.Thus,it is much helpful for companies to provide differentiated and personalized services for them.

关 键 词：K-MEANS算法 样本密度 客户细分 RFM模型 

分 类 号：TP311.13[自动化与计算机技术—计算机软件与理论] F274[自动化与计算机技术—计算机科学与技术]