Saved in:
Bibliographic Details
Main Authors: Demano, Joël Tatang, Dobson, Paul, Zygalakis, Konstantinos
Format: Preprint
Published: 2025
Subjects:
Online Access:https://arxiv.org/abs/2508.05462
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910178238005248
author Demano, Joël Tatang
Dobson, Paul
Zygalakis, Konstantinos
author_facet Demano, Joël Tatang
Dobson, Paul
Zygalakis, Konstantinos
contents In this paper, we propose a novel class of Piecewise Deterministic Markov Processes (PDMPs) that are designed to sample from probability distributions $π$ supported on a convex set $\mathcal{M}$. This class of PDMPs adapts the concept of a mirror map from convex optimisation to address sampling problems. The corresponding algorithms provide unbiased samples that respect the constraints and, moreover, allow for exact subsampling. We demonstrate the advantages of these algorithms against a range of constrained sampling problems where the proposed algorithms outperform state of the art stochastic differential equation-based methods.
format Preprint
id arxiv_https___arxiv_org_abs_2508_05462
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Piecewise Deterministic Sampling for Constrained Distributions
Demano, Joël Tatang
Dobson, Paul
Zygalakis, Konstantinos
Computation
Probability
60J25, 90C25, 60J22
In this paper, we propose a novel class of Piecewise Deterministic Markov Processes (PDMPs) that are designed to sample from probability distributions $π$ supported on a convex set $\mathcal{M}$. This class of PDMPs adapts the concept of a mirror map from convex optimisation to address sampling problems. The corresponding algorithms provide unbiased samples that respect the constraints and, moreover, allow for exact subsampling. We demonstrate the advantages of these algorithms against a range of constrained sampling problems where the proposed algorithms outperform state of the art stochastic differential equation-based methods.
title Piecewise Deterministic Sampling for Constrained Distributions
topic Computation
Probability
60J25, 90C25, 60J22
url https://arxiv.org/abs/2508.05462