Low Stein Discrepancy via Message-Passing Monte Carlo

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Kirk, Nathan, Rusch, T. Konstantin, Zech, Jakob, Rus, Daniela
Format: Preprint
Published: 2025
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913761022967808
author Kirk, Nathan
Rusch, T. Konstantin
Zech, Jakob
Rus, Daniela
author_facet Kirk, Nathan
Rusch, T. Konstantin
Zech, Jakob
Rus, Daniela
contents Message-Passing Monte Carlo (MPMC) was recently introduced as a novel low-discrepancy sampling approach leveraging tools from geometric deep learning. While originally designed for generating uniform point sets, we extend this framework to sample from general multivariate probability distributions with known probability density function. Our proposed method, Stein-Message-Passing Monte Carlo (Stein-MPMC), minimizes a kernelized Stein discrepancy, ensuring improved sample quality. Finally, we show that Stein-MPMC outperforms competing methods, such as Stein Variational Gradient Descent and (greedy) Stein Points, by achieving a lower Stein discrepancy.
format Preprint
id arxiv_https___arxiv_org_abs_2503_21103
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Low Stein Discrepancy via Message-Passing Monte Carlo
Kirk, Nathan
Rusch, T. Konstantin
Zech, Jakob
Rus, Daniela
Machine Learning
Numerical Analysis
Message-Passing Monte Carlo (MPMC) was recently introduced as a novel low-discrepancy sampling approach leveraging tools from geometric deep learning. While originally designed for generating uniform point sets, we extend this framework to sample from general multivariate probability distributions with known probability density function. Our proposed method, Stein-Message-Passing Monte Carlo (Stein-MPMC), minimizes a kernelized Stein discrepancy, ensuring improved sample quality. Finally, we show that Stein-MPMC outperforms competing methods, such as Stein Variational Gradient Descent and (greedy) Stein Points, by achieving a lower Stein discrepancy.
title Low Stein Discrepancy via Message-Passing Monte Carlo
topic Machine Learning
Numerical Analysis
url https://arxiv.org/abs/2503.21103