基于属性加权的RCM算法  

A RCM algorithm based on feature weights

作　　者：张朋[1] 戴月明[1] 吴定会[1] ZHANG Peng;DAI Yue-ming;WU Ding-hui(School of Internet of Things Engineering,Jiangnan University,Wuxi 214122,China)

机构地区：[1]江南大学物联网工程学院,江苏无锡214122

出　　处：《计算机工程与科学》2018年第3期544-549,共6页Computer Engineering & Science

基　　金：国家863计划(2013AA040405)

摘　　要：传统的粗糙集均值算法RCM的聚类准则是建立在参与聚类的属性同等重要的假设下,而在自然场景下的聚类问题中,不同的属性对聚类结果的影响是不同的。针对该问题,提出了将聚类属性进行加权处理的WRCM算法。具体地,为了筛选出对聚类结果产生关键影响的具有辨别力的聚类属性,算法通过引入权重矩阵将不同的属性赋予不同的属性权重。实验结果表明,本算法可以达到属性选择的效果,从而提高了最终的聚类精确度。The clustering criterion of the conventional Rough C-Means(RCM)algorithm is based on the hy-pothesis that the attributes involved in clustering are equally important.However,in the natural scenario cluste-ring problems,the different attributes have different effects on the clustering results.To address this issue,we propose a weighted RCM by weighting clustering attributes.Specifically,in order to filter out discerning cluste-ring attributes that have a crucial impact on the clustering results,the algorithm assigns different attributes to different attribute weights by introducing a weight matrix.The experimental results show that the proposed method is able to extract the attributes,which improves the clustering accuracy.

关 键 词：粗糙集均值算法 聚类 关键属性 权重矩阵 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]