Orthogonal MCMC algorithms

Luca Martino, Víctor Elvira, David Luengo, Antonio Artés-Rodríguez, Jukka Corander

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract

Monte Carlo (MC) methods are widely used in signal processing, machine learning and stochastic optimization. A well-known class of MC methods are Markov Chain Monte Carlo (MCMC) algorithms. In this work, we introduce a novel parallel interacting MCMC scheme, where the parallel chains share information using another MCMC technique working on the entire population of current states. These parallel "vertical" chains are led by random-walk proposals, whereas the "horizontal" MCMC uses a independent proposal, which can be easily adapted by making use of all the generated samples. Numerical results show the advantages of the proposed sampling scheme in terms of mean absolute error, as well as robustness w.r.t. to initial values and parameter choice.

Original languageEnglish
Title of host publication2014 IEEE Workshop on Statistical Signal Processing, SSP 2014
PublisherIEEE Computer Society
Pages364-367
Number of pages4
ISBN (Print)9781479949755
DOIs
Publication statusPublished - 1 Jan 2014
Event2014 IEEE Workshop on Statistical Signal Processing, SSP 2014 - Gold Coast, QLD, Australia
Duration: 29 Jun 20142 Jul 2014

Conference

Conference2014 IEEE Workshop on Statistical Signal Processing, SSP 2014
CountryAustralia
CityGold Coast, QLD
Period29/06/142/07/14

Keywords

  • Bayesian inference
  • Markov Chain Monte Carlo (MCMC)
  • Parallel Chains
  • Population Monte Carlo

Fingerprint Dive into the research topics of 'Orthogonal MCMC algorithms'. Together they form a unique fingerprint.

Cite this