Paper: Sometimes Average is Best: The Importance of Averaging for Prediction using MCMC Inference in Topic Modeling

ACL ID D14-1182
Title Sometimes Average is Best: The Importance of Averaging for Prediction using MCMC Inference in Topic Modeling
Venue Conference on Empirical Methods in Natural Language Processing
Session Main Conference
Year 2014
Authors

Markov chain Monte Carlo (MCMC) approxi- mates the posterior distribution of latent vari- able models by generating many samples and averaging over them. In practice, however, it is often more convenient to cut corners, using only a single sample or following a suboptimal averaging strategy. We systematically study dif- ferent strategies for averaging MCMC samples and show empirically that averaging properly leads to significant improvements in prediction.