High-accuracy sampling from constrained spaces with the Metropolis-adjusted Preconditioned Langevin Algorithm
Abstract
In this work, we propose a first-order sampling method called the Metropolis-adjusted Preconditioned Langevin Algorithm for approximate sampling from a target distribution whose support is a proper convex subset of . Our proposed method is the result of applying a Metropolis-Hastings filter to the Markov chain formed by a single step of the preconditioned Langevin algorithm with a metric , and is motivated by the natural gradient descent algorithm for optimisation. We derive non-asymptotic upper bounds for the mixing time of this method for sampling from target distributions whose potentials are bounded relative to , and for exponential distributions restricted to the support. Our analysis suggests that if satisfies stronger notions of self-concordance introduced in Kook and Vempala (2024), then these mixing time upper bounds have a strictly better dependence on the dimension than when is merely self-concordant. We also provide numerical experiments that demonstrates the practicality of our proposed method. Our method is a high-accuracy sampler due to the polylogarithmic dependence on the error tolerance in our mixing time upper bounds.
Cite
@article{arxiv.2412.18701,
title = {High-accuracy sampling from constrained spaces with the Metropolis-adjusted Preconditioned Langevin Algorithm},
author = {Vishwak Srinivasan and Andre Wibisono and Ashia Wilson},
journal= {arXiv preprint arXiv:2412.18701},
year = {2025}
}
Comments
55 pages, 5 figures, 2 tables. Shorter version without experiments accepted at ALT 2025. v3: fixes minor typographical errors