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.

The Wigner Branching Random Walk: Efficient Implementation and Performance Evaluation

Efficient implementation and performance evaluation of the Wigner branching random walk

A Branching Random Walk Method for Many-Body Wigner Quantum Dynamics

Branching Random Walk Solutions to the Wigner Equation

Comparison of deterministic and stochastic methods for time-dependent Wigner simulations

Overcoming the numerical sign problem in the Wigner dynamics via adaptive particle annihilation

An Advective-Spectral-Mixed Method for Time-Dependent Many-Body Wigner Simulations.

Simulating rare events using a Weighted Ensemble-based string method

Optimization for the propagation of a multiparticle quantum walk in a one-dimensional lattice

Dynamically Optimized Wang-Landau Sampling with Adaptive Trial Moves and Modification Factors

Self-repellent branching random walk

The Bose-Einstein effect in Monte-Carlo generators: weight methods

On importance sampling and independent Metropolis-Hastings with an unbounded weight function

A stochastic algorithm approach for the elephant random walk with applications

Weighted Jump in Random Walk graph sampling

Branching random walks in random environment: a survey

Numerical implementation of some reweighted path integral methods

Weighted ensemble: Recent mathematical developments

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

Reweight-annealing method for evaluating the partition function via quantum Monte Carlo calculations

The Brownian motion, Wiener process, and Pu migration as examples of random walk