Data Mining and Knowledge Discovery

, Volume 27, Issue 2, pp 193–224

How to “alternatize” a clustering algorithm

Authors

    • Department of Mathematics and Computer ScienceVirginia State University
  • Naren Ramakrishnan
    • Department of Computer ScienceVirginia Polytechnic Institute and State University
  • Ian Davidson
    • Department of Computer ScienceUniversity of California
  • Layne T. Watson
    • Department of Computer ScienceVirginia Polytechnic Institute and State University
    • Department of MathematicsVirginia Polytechnic Institute and State University
Article

DOI: 10.1007/s10618-012-0288-4

Cite this article as:
Hossain, M.S., Ramakrishnan, N., Davidson, I. et al. Data Min Knowl Disc (2013) 27: 193. doi:10.1007/s10618-012-0288-4

Abstract

Given a clustering algorithm, how can we adapt it to find multiple, nonredundant, high-quality clusterings? We focus on algorithms based on vector quantization and describe a framework for automatic ‘alternatization’ of such algorithms. Our framework works in both simultaneous and sequential learning formulations and can mine an arbitrary number of alternative clusterings. We demonstrate its applicability to various clustering algorithms—k-means, spectral clustering, constrained clustering, and co-clustering—and effectiveness in mining a variety of datasets.

Keywords

ClusteringAlternative clustering

Copyright information

© The Author(s) 2012