一种基于加权复杂网络特征的K-means聚类算法  

A Novel K-means Clustering Algorithm Based on Weighted Complex Networks Feature

作　　者：赵鹏[1] 耿焕同[2] 蔡庆生[2] 王清毅[2] 

机构地区：[1]安徽大学计算智能与信号处理教育部重点实验室,安徽合肥230039 [2]中国科学技术大学计算机系,安徽合肥230027

出　　处：《计算机技术与发展》2007年第9期35-37,40,共4页Computer Technology and Development

基　　金：国家自然科学基金项目(70171052);安徽省高校青年教师基金项目(2006jq1040)

摘　　要：在分析了传统的基于划分的K-means聚类算法的优越性和存在不足的基础上,根据近两年复杂网络研究中部分新的理论成果,提出了复杂网络加权度、加权聚集度与加权聚集系数的定义,并将数据聚类转换为复杂网络上的节点聚类,提出基于加权复杂网络特征的K-means聚类算法(简称WCNFC算法)。实验结果表明,该算法根据节点加权复杂网络特征值,能够较好地找到聚类中心,有效地避免了对初始化选值敏感性的问题,从而使得聚类质量大大提高。After analyzing the advantages and disadvantages of the traditional partitioned K - means clustering algorithm and based on the new theory results achieved in the field of complex networks, the definitions of weighted degree, weighted clustering degree, and weighted clustering coefficient of complex networks and a novel K - means clustering algorithm based on the weighted complex networks feature were proposed. The clustering of datum was transformed into clustering of nodes in complex networks. The experimental results show that this algorithm can find clustering centers better based on the weighted complex networks feature of nodes and it is robust to initialization, so the quality of clustering is improved greatly.

关 键 词：聚类 复杂网络 聚集度 聚集系数 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]