Kinetic Langevin Splitting Schemes for Constrained Sampling

View PDF HTML (experimental)

Abstract:Constrained sampling is an important and challenging task in computational statistics, concerned with generating samples from a distribution under certain constraints. There are numerous types of algorithm aimed at this task, ranging from general Markov chain Monte Carlo, to unadjusted Langevin methods. In this article we propose a series of new sampling algorithms based on the latter of these, specifically the kinetic Langevin dynamics. Our series of algorithms are motivated on advanced numerical methods which are splitting order schemes, which include the BU and BAO families of splitting this http URL advantage lies in the fact that they have favorable strong order (bias) rates and computationally efficiency. In particular we provide a number of theoretical insights which include a Wasserstein contraction and convergence results. We are able to demonstrate favorable results, such as improved complexity bounds over existing non-splitting methodologies. Our results are verified through numerical experiments on a range of models with constraints, which include a toy example and Bayesian linear regression.

Comments: 35 pages

Subjects:

Methodology (stat.ME); Numerical Analysis (math.NA); Machine Learning (stat.ML)

Cite as: arXiv:2603.23397 [stat.ME]

(or arXiv:2603.23397v1 [stat.ME] for this version)

https://doi.org/10.48550/arXiv.2603.23397

arXiv-issued DOI via DataCite

Submission history

From: Neil Chada [view email] [v1] Tue, 24 Mar 2026 16:34:38 UTC (689 KB) [v2] Tue, 31 Mar 2026 05:21:41 UTC (689 KB)