Parallel tempering with a variational reference
Nov 28, 2022
Speakers
Nikola Surjanovic
Speaker · 0 followers
Saifuddin Syed
Speaker · 0 followers
Alexandre Bouchard-Côté
Speaker · 0 followers
Trevor Campbell
Speaker · 0 followers
About
Sampling from complex target distributions is a challenging task fundamental to Bayesian inference. Parallel tempering (PT) addresses this problem by constructing a Markov chain on the expanded state space of a sequence of distributions interpolating between the posterior distribution and a fixed reference distribution, which is typically chosen to be the prior. However, in the typical case where the prior and posterior are nearly mutually singular, PT methods are computationally prohibitive. In this work we address this challenge by constructing a generalized annealing path connecting the posterior to an adaptively tuned variational reference. The reference distribution is tuned to minimize the forward (inclusive) KL divergence to the posterior distribution using a simple, gradient-free moment-matching procedure. We show that our adaptive procedure converges to the forward KL minimizer, and that the forward KL divergence serves as a good proxy to a previously developed measure of PT performance. We also show that in the large-data limit in typical Bayesian models, the proposed method improves in performance, while traditional PT deteriorates arbitrarily. Finally, we introduce PT with two references—one fixed, one variational—with a novel split annealing path that ensures stable variational reference adaptation. The paper concludes with experiments that demonstrate the large empirical gains achieved by our method in a wide range of realistic Bayesian inference scenarios.Sampling from complex target distributions is a challenging task fundamental to Bayesian inference. Parallel tempering (PT) addresses this problem by constructing a Markov chain on the expanded state space of a sequence of distributions interpolating between the posterior distribution and a fixed reference distribution, which is typically chosen to be the prior. However, in the typical case where the prior and posterior are nearly mutually singular, PT methods are computationally prohibitive. In…
Organizer
NeurIPS 2022
Account · 952 followers
Like the format? Trust SlidesLive to capture your next event!
Professional recording and live streaming, delivered globally.
Sharing
Recommended Videos
Presentations on similar topic, category or speaker
Active Bayesian Causal Inference: A Bayesian Active Learning Framework for Integrated Causal Discovery and Reasoning
Watch later
Total of 0 viewers voted for saving the presentation to eternal vault which is 0.0%
Non-Linear Coordination Graphs
Watch later
Yipeng Kang, …
Total of 0 viewers voted for saving the presentation to eternal vault which is 0.0%
Contrastive Value Learning: Implicit Models for Simple Offline RL
Watch later
Total of 0 viewers voted for saving the presentation to eternal vault which is 0.0%
Extrapolation via Interpolation
Watch later
Total of 0 viewers voted for saving the presentation to eternal vault which is 0.0%
Tree Mover's Distance: Bridging Graph Metrics and Stability of Graph Neural Networks
Watch later
Total of 0 viewers voted for saving the presentation to eternal vault which is 0.0%
The Minority Matters: A Diversity-Promoting Collaborative Metric Learning Algorithm
Watch later
Shilong Bao, …
Total of 0 viewers voted for saving the presentation to eternal vault which is 0.0%