Abstract

Although the study of clustering is centered around an intuitively compelling goal, it has been very di#cult to develop a unified framework for reasoning about it at a technical level, and profoundly diverse approaches to clustering abound in the research community. Here we suggest a formal perspective on the difficulty in finding such a unification, in the form of an impossibility theorem: for a set of three simple properties, we show that there is no clustering function satisfying all three. Relaxations of these properties expose some of the interesting (and unavoidable) trade-offs at work in well-studied clustering techniques such as single-linkage, sum-of-pairs, k-means, and k-median.

Citations

1753	Dubes, Algorithms for Clustering Data - Jain, C - 1988
1196	An Introduction to Game Theory - Osborne - 2004
525	Cluster Analysis for Applications - Anderberg - 1973
448	Social Choice and Individual Values - Arrow - 1970
170	D: Pattern classification. 2nd edition - Duda, Hart, et al. - 2001
116	Discrete Location Theory - Mirchandani, Francis - 1990
51	Social choice theory and recommender systems: Analysis of the axiomatic foundations of collaborative filtering - Pennock, Horvitz, et al.
28	A theory of proximity based clustering: Structure detection by optimization. Pattern Recognition, 2000. Sabine Schulte im Walde. Experiments on the automatic induction of german semantic verb classes - Puzicha, Hofmann, et al.
9	private communication - Kalai - 2007
2	Approximation algorithms for geometric prolems,” in Approximation Algorithms for NP-Hard - Bern, Eppstein - 1996
2	An impossibility result in axiomatic location theory - Hansen, Roberts - 1996