Título: Consensus Clustering Using Weighted Association
Autores: Nanda, Aparajita; Pujari, Arun K.
Resumo: Consensus clustering has emerged as a method of improving quality and robustness in clustering by optimally combining the results of different clustering process. In last few years, several approaches are proposed. In this paper, we propose a new method of arriving at a consensus clustering. We assign confidence score to each partition in the ensemble and compute weighted co-association for all pairs of data objects. In order to derive the consensus clustering from the co-association matrix, we use cross-association technique to group the rows and columns simultaneously. The objective is to derive as many clusters of homogenous blocks as possible. The set of non-zero blocks are taken as the resulting partition. The use of cross-association technique captures the transitive relationship. We show empirically that for the benchmark datasets, our technique yields better consensus clustering than any other known algorithms.
Palavras-chave: Clustering ensemble; Co-association; Cross-association
Código DOI: 10.21528/CBIC2011-10.1
Artigo em pdf: st_10.1.pdf
Arquivo BibTex: st_10.1.bib