Rapid Convergence of the Unadjusted Langevin Algorithm: Isoperimetry Suffices
Abstract
We study the Unadjusted Langevin Algorithm (ULA) for sampling from a probability distribution on . We prove a convergence guarantee in Kullback-Leibler (KL) divergence assuming satisfies a log-Sobolev inequality and the Hessian of is bounded. Notably, we do not assume convexity or bounds on higher derivatives. We also prove convergence guarantees in R\'enyi divergence of order assuming the limit of ULA satisfies either the log-Sobolev or Poincar\'e inequality. We also prove a bound on the bias of the limiting distribution of ULA assuming third-order smoothness of , without requiring isoperimetry.
Keywords
Cite
@article{arxiv.1903.08568,
title = {Rapid Convergence of the Unadjusted Langevin Algorithm: Isoperimetry Suffices},
author = {Santosh S. Vempala and Andre Wibisono},
journal= {arXiv preprint arXiv:1903.08568},
year = {2022}
}
Comments
v4: Updated discussion and added properties of biased limit v3: Simplified analysis of R\'enyi divergence, improved exposition, and added figures v2: Added analysis of R\'enyi divergence and Poincar\'e assumption