基于K-means算法的电子商务客户细分研究

摘要

文章将数据挖掘技术引入电子商务领域的客户细分研究,为了从客户信息中挖掘出潜在的信息,并对客户进行分类管理,本文采用了聚类算法中基于半监督近邻传播的K-means算法运用于某服装电子商务网站进行客户细分,并详细介绍了新算法的改进过程、实现流程以及通过对某服装电子商务网站数据信息的具体试验,验证其算法改进后的有效优势,具体表现在对数据集中的噪声点能进行有效的排除,并能获取较为精准的初始聚类中心,以提高聚类质量,同时极大的提升了聚类的准确率和各聚类之间的紧密度.%In this paper,data mining technology is introduced into the customer segmentation of electronic commerce industry,in order to dig out the potential information from the customer information.,and to obtain the customer classification management,this paper adopts the K-means algorithm based on a semi-supervised neighbor spread from the clustering algorithm , applied to a clothing e-commerce site,and introduces the improvement process of new algorithm,the implementation process, and verifies the effective advantages about the improved algorithm, through the clothing e-commerce site's data information, embodied that the data set can effectively eliminate the noise of the point, and obtain more accurate initial clustering center, in order to improve the quality of clustering, at the same time, greatly improve the accuracy of clustering and tightness between each cluster.