Universality of the Langevin diffusion as scaling limit of a family of Metropolis-Hastings processes

07/24/2019

∙

Given a target distribution μ on a general state space X and a proposal Markov jump process with generator Q, in this paper we study the scaling limit of two types of Metropolis-Hastings (MH) processes with generators M_1(Q,μ) and M_2(Q,μ) respectively. First, we motivate our study of M_2 by offering a geometric interpretation of M_1, M_2 and their convex combinations as L^1 minimizers between Q and the set of μ-reversible generators of Markov jump processes, extending the results by Billera and Diaconis '01. Specializing into the case of X = R^d along with a Gaussian proposal with vanishing variance and Gibbs target distribution, we prove that, upon appropriate scaling in time, the family of Markov jump processes corresponding to M_1, M_2 or their convex combinations all converge weakly to an universal Langevin diffusion.

READ FULL TEXT

Universality of the Langevin diffusion as scaling limit of a family of Metropolis-Hastings processes

Universality of the Langevin diffusion as scaling limit of a family of Metropolis-Hastings processes I: fixed dimension

Stationary Distribution Convergence of the Offered Waiting Processes in Heavy Traffic under General Patience Time Scaling

Markov substitute processes : a new model for linguistics and beyond

Systematic approaches to generate reversiblizations of Markov chains

Mixing convergence of LSE for supercritical Gaussian AR(2) processes using random scaling

Exact targeting of Gibbs distributions using velocity-jump processes

Convergence of Dirichlet Forms for MCMC Optimal Scaling with General Target Distributions on Large Graphs

Universality of the Langevin diffusion as scaling limit of a family of Metropolis-Hastings processes

Related Research

Universality of the Langevin diffusion as scaling limit of a family of Metropolis-Hastings processes I: fixed dimension

Stationary Distribution Convergence of the Offered Waiting Processes in Heavy Traffic under General Patience Time Scaling

Markov substitute processes : a new model for linguistics and beyond

Systematic approaches to generate reversiblizations of Markov chains

Mixing convergence of LSE for supercritical Gaussian AR(2) processes using random scaling

Exact targeting of Gibbs distributions using velocity-jump processes

Convergence of Dirichlet Forms for MCMC Optimal Scaling with General Target Distributions on Large Graphs