Problems in data analysis often require the unsupervised partitioning of a
dataset into clusters. Many methods exist for such partitioning but most ha
ve the weakness of being model-based (most assuming hyper-ellipsoidal clust
ers) or computationally infeasible in anything more than a three-dimensiona
l data space. We re-consider the notion of cluster analysis in information-
theoretic terms and show that minimisation of partition entropy can be used
to estimate the number and structure of probable data generators. (C) 2000
Pattern Recognition Society. Published by Elsevier Science Ltd. All rights
reserved.