Randomized Runge-Kutta-Nyström Methods for Unadjusted Hamiltonian and Kinetic Langevin Monte Carlo

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Bou-Rabee, Nawaf, Kleppe, Tore Selland
Format: Preprint
Published: 2023
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866910817654407168
author Bou-Rabee, Nawaf
Kleppe, Tore Selland
author_facet Bou-Rabee, Nawaf
Kleppe, Tore Selland
contents We introduce $5/2$- and $7/2$-order $L^2$-accurate randomized Runge-Kutta-Nyström methods, tailored for approximating Hamiltonian flows within non-reversible Markov chain Monte Carlo samplers, such as unadjusted Hamiltonian Monte Carlo and unadjusted kinetic Langevin Monte Carlo. We establish quantitative $5/2$-order $L^2$-accuracy upper bounds under gradient and Hessian Lipschitz assumptions on the potential energy function. The numerical experiments demonstrate the superior efficiency of the proposed unadjusted samplers on a variety of well-behaved, high-dimensional target distributions.
format Preprint
id arxiv_https___arxiv_org_abs_2310_07399
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Randomized Runge-Kutta-Nyström Methods for Unadjusted Hamiltonian and Kinetic Langevin Monte Carlo
Bou-Rabee, Nawaf
Kleppe, Tore Selland
Numerical Analysis
Probability
Computation
Methodology
Machine Learning
60J05 (Primary) 65C05, 65L05, 65P10 (Secondary)
We introduce $5/2$- and $7/2$-order $L^2$-accurate randomized Runge-Kutta-Nyström methods, tailored for approximating Hamiltonian flows within non-reversible Markov chain Monte Carlo samplers, such as unadjusted Hamiltonian Monte Carlo and unadjusted kinetic Langevin Monte Carlo. We establish quantitative $5/2$-order $L^2$-accuracy upper bounds under gradient and Hessian Lipschitz assumptions on the potential energy function. The numerical experiments demonstrate the superior efficiency of the proposed unadjusted samplers on a variety of well-behaved, high-dimensional target distributions.
title Randomized Runge-Kutta-Nyström Methods for Unadjusted Hamiltonian and Kinetic Langevin Monte Carlo
topic Numerical Analysis
Probability
Computation
Methodology
Machine Learning
60J05 (Primary) 65C05, 65L05, 65P10 (Secondary)
url https://arxiv.org/abs/2310.07399