Consensus clustering via pivotal methods
Leonardo Egidi, Roberta Pappadà, Francesco Pauli, Nicola Torelli · ArTS Archivio della ricerca di Trieste (University of Trieste https://www.units.it/) · 2019
We propose an approach to the cluster ensemble problem based on pivotal units extracted from a co-association matrix. It can be seen as a modified version of K-means method, which utilizes pivots for careful seeding. Different criteria for identifying the pivots are discussed, as well as preliminary results concerning the comparison with alternative ensemble methods.