Constructing sampling schemes via coupling: Markov semigroups and optimal transport
arXiv:1806.11026
Abstract
In this paper we develop a general framework for constructing and analysing coupled Markov chain Monte Carlo samplers, allowing for both (possibly degenerate) diffusion and piecewise deterministic Markov processes. For many performance criteria of interest, including the asymptotic variance, the task of finding efficient couplings can be phrased in terms of problems related to optimal transport theory. We investigate general structural properties, proving a singularity theorem that has both geometric and probabilistic interpretations. Moreover, we show that those problems can often be solved approximately and support our findings with numerical experiments. For the particular objective of estimating the variance of a Bayesian posterior, our analysis suggests using novel techniques in the spirit of antithetic variates. Addressing the convergence to equilibrium of coupled processes we furthermore derive a modified Poincaré inequality.
arXiv comment54 pages, 5 figures
References in corpus (11)
- Stein Variational Gradient Descent: A General Purpose Bayesian Inference Algorithm
- The computation of averages from equilibrium and nonequilibrium Langevin molecular dynamics
- Variance Reduction using Nonreversible Langevin Samplers
- Computational Optimal Transport
- Piecewise-Deterministic Markov Chain Monte Carlo
- Ergodicity of the zigzag process
- Multiprocess parallel antithetic coupling for backward and forward Markov Chain Monte Carlo
- Improving the convergence of reversible samplers
- Gibbs flow for approximate transport with applications to Bayesian computation
- Existence of (Markovian) solutions to martingale problems associated with Lévy-type operators
- Circularly-Coupled Markov Chain Sampling