PASCAL - Pattern Analysis, Statistical Modelling and Computational Learning

Collapsed Variational Inference for HDP
Yee Whye Teh, Kenichi Kurihara and Max Welling
In: NIPS 2007, 03 Dec - 08 Dec 2007, Vancouver, Canada.

Abstract

A wide variety of Dirichlet-multinomial `topic' models have found interesting applications in recent years. While Gibbs sampling remains an important method of inference in such models, variational techniques have certain advantages such as easy assessment of convergence, easy optimization without the need to maintain detailed balance, a bound on the marginal likelihood, and side-stepping of issues with topic-identifiability. The most accurate variational technique thus far, namely collapsed variational latent Dirichlet allocation, did not deal with model selection nor did it include inference for hyperparameters. We address both issues by generalizing the technique, obtaining the first variational algorithm to deal with the hierarchical Dirichlet process and to deal with hyperparameters of Dirichlet variables. Experiments show a significant improvement in accuracy.

EPrint Type:Conference or Workshop Item (Paper)
Project Keyword:Project Keyword UNSPECIFIED
Subjects:Learning/Statistics & Optimisation
Theory & Algorithms
Information Retrieval & Textual Information Access
ID Code:3791
Deposited By:Yee Whye Teh
Deposited On:25 February 2008