Collapsed Variational Dirichlet Process Mixture Models

Kenichi Kurihara, Max Welling, Yee Whye Teh

Nonparametric Bayesian mixture models, in particular Dirichlet process (DP) mixture models, have shown great promise for density estimation and data clustering. Given the size of today's datasets, computational efficiency becomes an essential ingredient in the applicability of these techniques to real world data. We study and experimentally compare a number of variational Bayesian (VB) approximations to the DP mixture model. In particular we consider the standard VB approximation where parameters are assumed to be independent from cluster assignment variables, and a novel collapsed VB approximation where mixture weights are marginalized out. For both VB approximations we consider two different ways to approximate the DP, by truncating the stick-breaking construction, and by using a finite mixture model with a symmetric Dirichlet prior.

Subjects: 12. Machine Learning and Discovery


Submitted: Oct 16, 2006

This page is copyrighted by AAAI. All rights reserved. Your use of this site constitutes acceptance of all of AAAI's terms and conditions and privacy policy.