Online but Accurate Inference for Latent Variable Models with Local Gibbs Sampling
This work addresses the problem of efficient and accurate inference for researchers and practitioners using latent variable models, offering incremental improvements over prior methods.
The paper tackles parameter inference in large-scale latent variable models by proposing a novel online inference method using local Gibbs sampling, which outperforms existing methods in experiments on latent Dirichlet allocation, achieving superior test log-likelihoods.
We study parameter inference in large-scale latent variable models. We first propose an unified treatment of online inference for latent variable models from a non-canonical exponential family, and draw explicit links between several previously proposed frequentist or Bayesian methods. We then propose a novel inference method for the frequentist estimation of parameters, that adapts MCMC methods to online inference of latent variable models with the proper use of local Gibbs sampling. Then, for latent Dirich-let allocation,we provide an extensive set of experiments and comparisons with existing work, where our new approach outperforms all previously proposed methods. In particular, using Gibbs sampling for latent variable inference is superior to variational inference in terms of test log-likelihoods. Moreover, Bayesian inference through variational methods perform poorly, sometimes leading to worse fits with latent variables of higher dimensionality.