Handbook of Cluster Analysis (provisional top level file)

This chapter describes the problem of combining multiple partitionings of a set of objects into a single consolidated clustering without accessing the features or algorithms that determine these partitionings – popularly known as the problem of “consensus clustering”. We illustrate different algorithms for solving the consensus clustering problem. The notion of dissimilarity between a pair of clustering solutions plays a key role in designing any cluster ensemble algorithm and a summary of such dissimilarity measures is also provided. We also cover recent efforts on combining classifier and clustering ensembles, leading to new approaches for semi-supervised learning and transfer learning. Finally, we describe several applications of consensus clustering.