基于密度的购物篮数据聚类方法  

Density-based clustering for market basket data

作　　者：张平庸[1] 欧阳为民[1] 万志华[1] 

机构地区：[1]上海大学计算机科学与工程系,上海200072

出　　处：《计算机工程与设计》2005年第1期180-181,184,共3页Computer Engineering and Design

摘　　要：在现有的算法DBSCAN基础上,提出一种基于密度的处理购物篮事务数据的聚类方法-DCMBD(density-based clustering for market basket data)。使用了一种新的事务表示法,解决了购物篮数据的高维性和稀疏性问题。并对算法进行 了相应的改进,从而提高了聚类速度。实验结果表明此方法是有效可行的。A scheme of density-based clustering for market basket data is proposed. A new representation method is developed, which, in contrast to existing approaches, is proven to provide correct results. New algorithms are proposed for the processing of clustering. The scheme resolves the problem of high dimension and large sparseness in basket data and improves efficiency of clustering. Experimental results show the feasibility and effectiveness of the proposed scheme.

关 键 词：聚类 购物篮数据 数据挖掘 算法DBSCAN 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]