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.

A Particle-Inspired Monte Carlo Tree Estimation Method in Bayesian Filtering

Optimal Sampling Function-Based Particle Filter and Bayesian Estimation

Independent Particle Filters

A Novel Particle Filter Based on Hybrid Deterministic and Random Sampling

Estimation with particle filter under model uncertainty

Particle Gibbs for Bayesian Additive Regression Trees

An Improved Particle Filter Based on Space Distribution and Time Series Analysis

The particle filter algorithm based on evolution sampling

Improved Particle Filter Based on Fine Resampling Algorithm

A Post-Resampling Based Particle Filter for Online Bayesian Estimation and Tracking

Adaptive Particle Filter With Randomized Quasi-Monte Carlo Sampling for Unbalanced Distribution System State Estimation

On sequential Monte Carlo sampling methods for Bayesian filtering

Divide, Conquer, Combine Bayesian Decision Tree Sampling

Research of particle filter algorithm based on randomized Halton sequences

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

Particle Filter with Importance Density Function Generated by Updated System Equation

Uniform convergence over time of a nested particle filtering scheme for recursive parameter estimation in state--space Markov models

Deterministic resampling: Unbiased sampling to avoid sample impoverishment in particle filters

The Alive Particle Filter and Its Use in Particle Markov Chain Monte Carlo

A simple modification for improving inference of non-linear dynamical systems

A Discrete Bouncy Particle Sampler