%0 Report
%9 Technical Report
%A Kamvar, Sepandar D.
%A Klein, Dan
%A Manning, Christopher D.
%D 2002
%F ilprints:529
%I Stanford InfoLab
%K clustering, probabilistic models, model-based clustering, hierarchical clustering
%T Interpreting and Extending Classical Agglomerative Clustering Algorithms using a Model-Based Approach
%U http://ilpubs.stanford.edu:8090/529/
%X We present two results which arise from a model-based approach to hierarchical agglomerative clustering.  First, we show formally that the common heuristic agglomerative clustering algorithms -- single-link, complete-link, group-average, and Ward's method -- are each equivalent to a hierarchical model-based method.  This interpretation gives a theoretical explanation of the empirical behavior of these algorithms, as well as a principled approach to resolving practical issues, such as number of clusters or the choice of method. Second, we show how a model-based approach can be used to extend these basic agglomerative algorithms.  We introduce adjusted complete-link, Mahalanobis-link, and line-link as variants of the classical agglomerative methods, and demonstrate their utility.