A cost function for similarity-based hierarchical clustering

10/16/2015

∙

by   Sanjoy Dasgupta, et al.

∙

0

∙

share

The development of algorithms for hierarchical clustering has been hampered by a shortage of precise objective functions. To help address this situation, we introduce a simple cost function on hierarchies over a set of points, given pairwise similarities between those points. We show that this criterion behaves sensibly in canonical instances and that it admits a top-down construction procedure with a provably good approximation ratio.