Saved in:
Bibliographic Details
Main Authors: Hird, Max, Maire, Florian
Format: Preprint
Published: 2024
Subjects:
Online Access:https://arxiv.org/abs/2411.11983
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866913580288311296
author Hird, Max
Maire, Florian
author_facet Hird, Max
Maire, Florian
contents Autocorrelations in MCMC chains increase the variance of the estimators they produce. We propose the occlusion process to mitigate this problem. It is a process that sits upon an existing MCMC sampler, and occasionally replaces its samples with ones that are decorrelated from the chain. We show that this process inherits many desirable properties from the underlying MCMC sampler, such as a Law of Large Numbers, convergence in a normed function space, and geometric ergodicity, to name a few. We show how to simulate the occlusion process at no additional time-complexity to the underlying MCMC chain. This requires a threaded computer, and a variational approximation to the target distribution. We demonstrate empirically the occlusion process' decorrelation and variance reduction capabilities on two target distributions. The first is a bimodal Gaussian mixture model in 1d and 100d. The second is the Ising model on an arbitrary graph, for which we propose a novel variational distribution.
format Preprint
id arxiv_https___arxiv_org_abs_2411_11983
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle The occlusion process: improving sampler performance with parallel computation and variational approximation
Hird, Max
Maire, Florian
Computation
Autocorrelations in MCMC chains increase the variance of the estimators they produce. We propose the occlusion process to mitigate this problem. It is a process that sits upon an existing MCMC sampler, and occasionally replaces its samples with ones that are decorrelated from the chain. We show that this process inherits many desirable properties from the underlying MCMC sampler, such as a Law of Large Numbers, convergence in a normed function space, and geometric ergodicity, to name a few. We show how to simulate the occlusion process at no additional time-complexity to the underlying MCMC chain. This requires a threaded computer, and a variational approximation to the target distribution. We demonstrate empirically the occlusion process' decorrelation and variance reduction capabilities on two target distributions. The first is a bimodal Gaussian mixture model in 1d and 100d. The second is the Ising model on an arbitrary graph, for which we propose a novel variational distribution.
title The occlusion process: improving sampler performance with parallel computation and variational approximation
topic Computation
url https://arxiv.org/abs/2411.11983