Smelly parallel MCMC chains

L. Martino; V. Elvira; D. Luengo; A. Artés-Rodríguez; J. Corander

doi:10.1109/ICASSP.2015.7178736

Smelly parallel MCMC chains

L. Martino, V. Elvira, D. Luengo, A. Artés-Rodríguez, J. Corander

School of Mathematics

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract / Description of output

Monte Carlo (MC) methods are useful tools for Bayesian inference and stochastic optimization that have been widely applied in signal processing and machine learning. 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, thus yielding a faster exploration of the state space. The interaction is carried out generating a dynamic repulsion among the 'smelly' parallel chains that takes into account the entire population of current states. The ergodicity of the scheme and its relationship with other sampling methods are discussed. Numerical results show the advantages of the proposed approach in terms of mean square error, robustness w.r.t. to initial values and parameter choice.

Original language	English
Title of host publication	2015 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2015 - Proceedings
Publisher	Institute of Electrical and Electronics Engineers
Pages	4070-4074
Number of pages	5
Volume	2015-August
ISBN (Electronic)	9781467369978
DOIs	https://doi.org/10.1109/ICASSP.2015.7178736
Publication status	Published - 1 Jan 2015
Event	40th IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2015 - Brisbane, Australia Duration: 19 Apr 2014 → 24 Apr 2014

Conference

Conference	40th IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2015
Country/Territory	Australia
City	Brisbane
Period	19/04/14 → 24/04/14

Keywords / Materials (for Non-textual outputs)

Bayesian inference
Markov Chain Monte Carlo
parallel and interacting chains

Access to Document

10.1109/ICASSP.2015.7178736

Cite this

@inproceedings{726949a264d1474686fcbe9b19ace20a,

title = "Smelly parallel MCMC chains",

abstract = "Monte Carlo (MC) methods are useful tools for Bayesian inference and stochastic optimization that have been widely applied in signal processing and machine learning. 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, thus yielding a faster exploration of the state space. The interaction is carried out generating a dynamic repulsion among the 'smelly' parallel chains that takes into account the entire population of current states. The ergodicity of the scheme and its relationship with other sampling methods are discussed. Numerical results show the advantages of the proposed approach in terms of mean square error, robustness w.r.t. to initial values and parameter choice.",

keywords = "Bayesian inference, Markov Chain Monte Carlo, parallel and interacting chains",

author = "L. Martino and V. Elvira and D. Luengo and A. Art{\'e}s-Rodr{\'i}guez and J. Corander",

year = "2015",

month = jan,

day = "1",

doi = "10.1109/ICASSP.2015.7178736",

language = "English",

volume = "2015-August",

pages = "4070--4074",

booktitle = "2015 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2015 - Proceedings",

publisher = "Institute of Electrical and Electronics Engineers",

address = "United States",

note = "40th IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2015 ; Conference date: 19-04-2014 Through 24-04-2014",

}

Martino, L, Elvira, V, Luengo, D, Artés-Rodríguez, A & Corander, J 2015, Smelly parallel MCMC chains. in 2015 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2015 - Proceedings. vol. 2015-August, 7178736, Institute of Electrical and Electronics Engineers, pp. 4070-4074, 40th IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2015, Brisbane, Australia, 19/04/14. https://doi.org/10.1109/ICASSP.2015.7178736

TY - GEN

T1 - Smelly parallel MCMC chains

AU - Martino, L.

AU - Elvira, V.

AU - Luengo, D.

AU - Artés-Rodríguez, A.

AU - Corander, J.

PY - 2015/1/1

Y1 - 2015/1/1

N2 - Monte Carlo (MC) methods are useful tools for Bayesian inference and stochastic optimization that have been widely applied in signal processing and machine learning. 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, thus yielding a faster exploration of the state space. The interaction is carried out generating a dynamic repulsion among the 'smelly' parallel chains that takes into account the entire population of current states. The ergodicity of the scheme and its relationship with other sampling methods are discussed. Numerical results show the advantages of the proposed approach in terms of mean square error, robustness w.r.t. to initial values and parameter choice.

AB - Monte Carlo (MC) methods are useful tools for Bayesian inference and stochastic optimization that have been widely applied in signal processing and machine learning. 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, thus yielding a faster exploration of the state space. The interaction is carried out generating a dynamic repulsion among the 'smelly' parallel chains that takes into account the entire population of current states. The ergodicity of the scheme and its relationship with other sampling methods are discussed. Numerical results show the advantages of the proposed approach in terms of mean square error, robustness w.r.t. to initial values and parameter choice.

KW - Bayesian inference

KW - Markov Chain Monte Carlo

KW - parallel and interacting chains

UR - http://www.scopus.com/inward/record.url?scp=84946048708&partnerID=8YFLogxK

U2 - 10.1109/ICASSP.2015.7178736

DO - 10.1109/ICASSP.2015.7178736

M3 - Conference contribution

AN - SCOPUS:84946048708

VL - 2015-August

SP - 4070

EP - 4074

BT - 2015 IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2015 - Proceedings

PB - Institute of Electrical and Electronics Engineers

T2 - 40th IEEE International Conference on Acoustics, Speech, and Signal Processing, ICASSP 2015

Y2 - 19 April 2014 through 24 April 2014

ER -

Smelly parallel MCMC chains

Abstract / Description of output

Conference

Keywords / Materials (for Non-textual outputs)

Access to Document

Other files and links

Fingerprint

Cite this