Multi-view clustering ensembles

Abstract

Multi-view clustering and clustering ensembles have become increasingly popular in recent years. Multi-view clustering employs the relationship between views to cluster data; clustering ensembles combine different component c1usterings to a better final partition. In this paper, we proposed the multi-view clustering ensembles which extend clustering ensembles to multi-view clustering. Experimental results show good performances of multi-view spectral clustering ensembles and multi-view kernel k-means clustering ensembles on real datasets.