Deterministic Gibbs Sampling via Ordinary Differential Equations

Max Welling, University of Amsterdam
Event time: 
Monday, March 15, 2021 - 10:00am
Zoom Meeting ID: 97670014308 See map
Event description: 

Abstract:  Deterministic dynamics is an essential part of many MCMC algorithms, e.g. Hybrid Monte Carlo or samplers utilizing normalizing flows. This paper presents a general construction of deterministic measure-preserving dynamics using autonomous ODEs and tools from differential geometry.  We show how Hybrid Monte Carlo and other deterministic samplers follow as special cases of our theory. We then demonstrate the utility of our approach by constructing a continuous non-sequential version of Gibbs sampling in terms of an ODE flow and extending it to discrete state spaces. We find that our deterministic samplers are more sample efficient than stochastic counterparts, even if the latter generate independent samples.

email Ofir Lindenbaum <> for info.

Event Type: 
Applied Mathematics