Transport map unadjusted Langevin algorithms
Preprint
Abstract
Langevin dynamics are widely used in sampling high-dimensional, non-Gaussian distributions whose densities are known up to a normalizing constant. In particular, there is strong interest in unadjusted Langevin algorithms (ULA), which directly discretize Langevin dynamics to estimate expectations over the target distribution. We study the use of transport maps that approximately normalize a target distribution as a way to precondition and accelerate the convergence of Langevin dynamics. We show that in continuous time, when a transport map is applied to Langevin dynamics, the result is a Riemannian manifold Langevin dynamics (RMLD) with metric defined by the transport map. This connection suggests more systematic ways of learning metrics, and also yields alternative discretizations of the RMLD described by the map, which we study. Moreover, we show that under certain conditions, when the transport map is used in conjunction with ULA, we can improve the geometric rate of convergence of the output process in the 2--Wasserstein distance. Illustrative numerical results complement our theoretical claims.Keywords
All Related Versions
This publication has 32 references indexed in Scilit:
- Analysis of Langevin Monte Carlo from Poincaré to Log-SobolevFoundations of Computational Mathematics, 2024
- On the Representation and Learning of Monotone Triangular Transport MapsFoundations of Computational Mathematics, 2023
- Rapid Convergence of the Unadjusted Langevin Algorithm: Isoperimetry Suffices2023
- A transport-based multifidelity preconditioner for Markov chain Monte CarloAdvances in Computational Mathematics, 2019
- Improving the Convergence of Reversible SamplersJournal of Statistical Physics, 2016
- Sampling via Measure Transport: An Introduction2016
- Analysis and Geometry of Markov Diffusion Operators2014
- Optimal Non-reversible Linear Drift for the Convergence to Equilibrium of a DiffusionJournal of Statistical Physics, 2013
- A Family of Nonparametric Density Estimation AlgorithmsCommunications on Pure and Applied Mathematics, 2012
- Stability of Partially Implicit Langevin Schemes and Their MCMC VariantsMethodology and Computing in Applied Probability, 2010