Stereographic Markov Chain Monte Carlo

05/24/2022
by   Jun Yang, et al.
0

High dimensional distributions, especially those with heavy tails, are notoriously difficult for off-the-shelf MCMC samplers: the combination of unbounded state spaces, diminishing gradient information, and local moves, results in empirically observed "stickiness" and poor theoretical mixing properties – lack of geometric ergodicity. In this paper, we introduce a new class of MCMC samplers that map the original high dimensional problem in Euclidean space onto a sphere and remedy these notorious mixing problems. In particular, we develop random-walk Metropolis type algorithms as well as versions of Bouncy Particle Sampler that are uniformly ergodic for a large class of light and heavy-tailed distributions and also empirically exhibit rapid convergence in high dimensions. In the best scenario, the proposed samplers can enjoy the “blessings of dimensionality” that the mixing time decreases with dimension.

READ FULL TEXT

page 1

page 2

page 3

page 4

research
03/28/2019

Approximate spectral gaps for Markov chains mixing times in high-dimension

This paper introduces a concept of approximate spectral gap to analyze t...
research
09/02/2019

Fourier transform MCMC, heavy tailed distributions and geometric ergodicity

Markov Chain Monte Carlo methods become increasingly popular in applied ...
research
05/12/2021

Dimension-free Mixing for High-dimensional Bayesian Variable Selection

Yang et al. (2016) proved that the symmetric random walk Metropolis–Hast...
research
08/07/2020

MCMC Algorithms for Posteriors on Matrix Spaces

We study Markov chain Monte Carlo (MCMC) algorithms for target distribut...
research
01/11/2021

Complexity analysis of Bayesian learning of high-dimensional DAG models and their equivalence classes

We consider MCMC methods for learning equivalence classes of sparse Gaus...
research
11/22/2022

A new and asymptotically optimally contracting coupling for the random walk Metropolis

The reflection-maximal coupling of the random walk Metropolis (RWM) algo...
research
09/05/2022

On free energy barriers in Gaussian priors and failure of MCMC for high-dimensional unimodal distributions

We exhibit examples of high-dimensional unimodal posterior distributions...

Please sign up or login with your details

Forgot password? Click here to reset