Partition Selection Approach for Hierarchical Clustering Based on Clustering Ensemble
Hierarchical clustering algorithms are widely used in many fields of investigation. They provide a hierarchy of partitions of the same dataset. However, in many practical problems, the selection of a representative level (partition) in the hierarchy is needed. The classical approach to do so is by using a cluster validity index to select the best partition according to the criterion imposed by this index. In this paper, we present a new approach based on the clustering ensemble philosophy. The representative level is defined here as the consensus partition in the hierarchy. In the consensus computation process, we take into account the similarity between partitions and information from the evaluation of partitions with different cluster validity indexes. An experimental comparison on several datasets shows the superiority of the proposed approach with respect to the classical approach.
KeywordsHierarchical clustering partition selection clustering ensemble cluster validity index
- 3.Xu, R., Wunsch, D.C.: Clustering. IEEE Press Series on Computational Intelligence. John Wiley & Sons, Chichester (2009)Google Scholar
- 8.Bakir, G., Weston, J., Scholkopf, B.: Learning to find pre-images. In: Thrun, S., Saul, L. (eds.) Advances in Neural Information Processing Systems (NIPS 2003), vol. 16, pp. 449–456. MIT Press, Cambridge (2004)Google Scholar