Non-Reversible Markov Chain Monte Carlo

Researchers: Sergio Bacallado, Sam Power, Torben Sell

Classical Monte Carlo sampling techniques have been devised by constructing a Markov Process which is reversible with respect to the desired target measure. In recent years, however, it has become clear that using _non-reversible_ processes to sample is i) theoretically preferable, and ii) practically feasible. The goal of this project is to i) understand the theoretical properties of existing non-reversible samplers (e.g. Piecewise-Deterministic Markov Processes, such as the Bouncy Particle Sampler), and ii) to devise new non-reversible sampling techniques for different application domains.

Who's involved


There's still time to register to attend the CCIMI Connecting with Industry Day tomorrow, starting at 12:45pm (GMT)… View on Twitter