基于加权co-occurrence矩阵的聚类集成算法

聚类集成是数据挖掘研究的一个热点。它是利用同一数据集的多个聚类划分集成在一起,以提高聚类分析的性能。当前相关研究大多没有考虑进行集成的聚类成员的质量,因此较差的成员会对集成结果产生不良影响。文中提出了一种基于加权co-occurrence矩阵的聚类集成算法(WCSCE)。该方法首先计算出聚类成员基于属性值的co-occurrence矩阵,然后对聚类成员的质量进行简单评价并赋予权重,生成加权co-occurrence矩阵,进而产生集成结果。最后通过实验验证了该算法的有效性,并提高了聚类质量。

Weighted Cluster Ensemble Based on Co-Occurrence Matrix

Cluster ensemble is a hot topic in data mining research.It can find a combined clustering with better quality from multiple partitions.Most of resent researches pay little attention to the qualities of cluster members.However,bad cluster members and noise may affect the ensemble result.This paper presents a clustering ensemble algorithm based on weighted co-occurrence matrix.First the co-occurrence property value matrix of the cluster members is calculated.The significance of each cluster member is evaluated through information measures of clustering evaluation.Then weighted co-occurrence matrix is generated and the final ensemble result is obtained.Experimental results show the effectiveness of the algorithms,and the clustering accuracy is improved.