[NO15] Weighted multiblock clustering

Conférence Internationale avec comité de lecture : CLADAG 2015 10th Scientific Meeting of the Classification and Data Analysis Group, October 8-10, 2015, October 2015, pp.pp, Santa Margherita di Pula (Cagliari), Italy,

Auteurs: N. Niang Keita , M. Ouattara

Mots clés: CONSENSUS CLUSTERING, MULTIBLOCK DATA, RV index

Résumé: We address the issue of clustering individuals described by several homogeneous blocks of variables. Reformulating it as a problem of consensus of partitions, we propose a weighted consensus method based on the RV index to find the partition of the individuals. The principal idea of this consensus methods is to agglomerate the separate initial partitions obtain from each block into a global partition which has to be the most similar to the initial partitions according to the RV coefficient. The proposed method is illustrated and compared to CSPA (cluster based similarity partitioning algorithm) and a weighted consensus clustering method based on non-negative matrix factorization.

BibTeX

@inproceedings {
	NO15,
	title	=	"{Weighted multiblock clustering}",
	author	=	" N. Niang Keita and M. Ouattara ",
	booktitle	=	"{ CLADAG 2015 10th Scientific Meeting of the Classification and Data Analysis Group, October 8-10, 2015}",
	year	=	2015,
	month	=	"October",
	pages	=	"pp",
	address	=	"Santa Margherita di Pula (Cagliari), Italy",
}