Abstract:Many Markov chain Monte Carlo techniques currently available rely on discrete-time reversible Markov processes whose transition kernels are variations of the Metropolis–Hastings algorithm. We explore and generalize an alternative scheme recently introduced in the physics literature (Peters and de With 2012) where the target distribution is explored using a continuous-time nonreversible piecewise-deterministic Markov process. In the Metropolis–Hastings algorithm, a trial move to a region of lower target density, equivalently of higher “energy,” than the current state can be rejected with positive probability. In this alternative approach, a particle moves along straight lines around the space and, when facing a high energy barrier, it is not rejected but its path is modified by bouncing against this barrier. By reformulating this algorithm using inhomogeneous Poisson processes, we exploit standard sampling techniques to simulate exactly this Markov process in a wide range of scenarios of interest. Additionally, when the target distribution is given by a product of factors dependent only on subsets of the state variables, such as the posterior distribution associated with a probabilistic graphical model, this method can be modified to take advantage of this structure by allowing computationally cheaper “local” bounces, which only involve the state variables associated with a factor, while the other state variables keep on evolving. In this context, by leveraging techniques from chemical kinetics, we propose several computationally efficient implementations. Experimentally, this new class of Markov chain Monte Carlo schemes compares favorably to state-of-the-art methods on various Bayesian inference tasks, including for high-dimensional models and large datasets. Supplementary materials for this article are available online.

Metropolis-adjusted interacting particle sampling

The Bouncy Particle Sampler: A Nonreversible Rejection-Free Markov Chain Monte Carlo Method

High-accuracy sampling from constrained spaces with the Metropolis-adjusted Preconditioned Langevin Algorithm

A Discrete Bouncy Particle Sampler

Fully-Asynchronous Distributed Metropolis Sampler with Optimal Speedup.

Piecewise deterministic sampling with splitting schemes

Bayesian sampling using interacting particles

Fast sampling from constrained spaces using the Metropolis-adjusted Mirror Langevin algorithm

Single-ensemble multilevel Monte Carlo for discrete interacting-particle methods

Adapting The Gibbs Sampler

On particle Gibbs sampling

Stochastic Gradient Piecewise Deterministic Monte Carlo Samplers

On Disturbance State-Space Models and the Particle Marginal Metropolis-Hastings Sampler

Group Importance Sampling for Particle Filtering and MCMC

Less interaction with forward models in Langevin dynamics

Sampling algorithms in statistical physics: a guide for statistics and machine learning

Efficient Irreversible Monte Carlo Samplers

Superposition Enhanced Nested Sampling

Auxiliary MCMC and particle Gibbs samplers for parallelisable inference in latent dynamical systems

Sampling metastable systems using collective variables and Jarzynski-Crooks paths

A note on the Metropolis-Hastings acceptance probabilities for mixture spaces