Cluster Analysis: Overview

This article discusses analysis techniques for clustering objects into hopefully meaningful sets. Hierarchical methods are presented for clustering both variables and cases. Examples are presented for variations on the methods. A nonhierarchical method, k-means, is discussed for continuous data. Finally, an example comparing the results of k-means clustering to an ad hoc method using principal components analysis is given. Keywords: clustering; distance functions; hierarchical; principal components; single-linkage