An O(log2 N) SMC2 Algorithm on Distributed Memory with an Approx. Optimal L-Kernel

Calibrating statistical models using Bayesian inference often requires both accurate and timely estimates of parameters of interest. Particle Markov Chain Monte Carlo (p-MCMC) and Sequential Monte Carlo Squared (SMC 2 ) are two methods that use an unbiased estimate of the log-likelihood obtained fro...

Full description

Saved in:
Bibliographic Details
Published in:IEEE/SICE/RSJ International Conference on Multisensor Fusion and Integration for Intelligent Systems pp. 1 - 8
Main Authors: Rosato, Conor, Varsi, Alessandro, Murphy, Joshua, Maskell, Simon
Format: Conference Proceeding
Language:English
Published: IEEE 27.11.2023
Subjects:
Bayes methods
Bayesian inference
biomedical applications
distributed memory
Markov processes
Monte Carlo methods
parallel algorithms
Parameter estimation
Particle filters
particle Markov Chain Monte Carlo
Program processors
Recycling
SMC 2
ISSN:2767-9357
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Abstract Calibrating statistical models using Bayesian inference often requires both accurate and timely estimates of parameters of interest. Particle Markov Chain Monte Carlo (p-MCMC) and Sequential Monte Carlo Squared (SMC 2 ) are two methods that use an unbiased estimate of the log-likelihood obtained from a particle filter (PF) to evaluate the target distribution. P-MCMC constructs a single Markov chain which is sequential by nature so cannot be readily parallelized using Distributed Memory (DM) architectures. This is in contrast to SMC 2 which includes processes, such as importance sampling, that are described as embarrassingly parallel. However, difficulties arise when attempting to parallelize resampling. None-the-less, the choice of backward kernel, recycling scheme and compatibility with DM architectures makes SMC 2 an attractive option when compared with p-MCMC. In this paper, we present an SMC 2 framework that includes the following features: an optimal (in terms of time complexity) \mathcal{O}(\log_2 N) parallelization for DM architectures, an approximately optimal (in terms of accuracy) backward kernel, and an efficient recycling scheme. On a cluster of 128 DM processors, the results on a biomedical application show that SMC 2 achieves up to a 70× speed-up vs its sequential implementation. It is also more accurate and roughly 54× faster than p-MCMC. A GitHub link is given which provides access to the code.
AbstractList Calibrating statistical models using Bayesian inference often requires both accurate and timely estimates of parameters of interest. Particle Markov Chain Monte Carlo (p-MCMC) and Sequential Monte Carlo Squared (SMC 2 ) are two methods that use an unbiased estimate of the log-likelihood obtained from a particle filter (PF) to evaluate the target distribution. P-MCMC constructs a single Markov chain which is sequential by nature so cannot be readily parallelized using Distributed Memory (DM) architectures. This is in contrast to SMC 2 which includes processes, such as importance sampling, that are described as embarrassingly parallel. However, difficulties arise when attempting to parallelize resampling. None-the-less, the choice of backward kernel, recycling scheme and compatibility with DM architectures makes SMC 2 an attractive option when compared with p-MCMC. In this paper, we present an SMC 2 framework that includes the following features: an optimal (in terms of time complexity) \mathcal{O}(\log_2 N) parallelization for DM architectures, an approximately optimal (in terms of accuracy) backward kernel, and an efficient recycling scheme. On a cluster of 128 DM processors, the results on a biomedical application show that SMC 2 achieves up to a 70× speed-up vs its sequential implementation. It is also more accurate and roughly 54× faster than p-MCMC. A GitHub link is given which provides access to the code.
Author Murphy, Joshua
Varsi, Alessandro
Rosato, Conor
Maskell, Simon
Author_xml – sequence: 1
  givenname: Conor
  surname: Rosato
  fullname: Rosato, Conor
  email: cmrosa@liverpool.ac.uk
  organization: University of Liverpool,Department of Pharmacology and Therapeutics,United Kingdom
– sequence: 2
  givenname: Alessandro
  surname: Varsi
  fullname: Varsi, Alessandro
  email: a.varsi@liverpool.ac.uk
  organization: University of Liverpool,Department of Electrical Engineering and Electronics,United Kingdom
– sequence: 3
  givenname: Joshua
  surname: Murphy
  fullname: Murphy, Joshua
  email: joshua.murphy@liverpool.ac.uk
  organization: University of Liverpool,Department of Electrical Engineering and Electronics,United Kingdom
– sequence: 4
  givenname: Simon
  surname: Maskell
  fullname: Maskell, Simon
  email: smaskell@liverpool.ac.uk
  organization: University of Liverpool,Department of Electrical Engineering and Electronics,United Kingdom
BookMark eNo1kE1Lw0AYhFdRsNb-Aw-LJz0k7rvvfh5Da7WYGLC9lzXd1EiahE1E--8NqKcZGJ5hmEty1rSNJ-QGWAzA7P16sYyy5UpaKWTMGccYGCoQkp-QmdXWoGRouDT6lEy4VjqyKPUFmfX9B2MMFIyompDXpKH5bd3uOX25o-tszmlS79tQDe8H2jZ0UfVDqN4-B7-jmT-04Ui_xoy6hiZdF9rvmObdUB1cTdPo2YfG11fkvHR172d_OiWb5cNm_hSl-eNqnqRRJSyLnEBtRFFohc6XBvTomHPOMFWWKL0pQJYgCoYFdwyd5OB2YFCUVoHWBU7J9W9t5b3fdmHcEI7b_xfwB0YlUXI
ContentType Conference Proceeding
DBID 6IE
6IL
CBEJK
RIE
RIL
DOI 10.1109/SDF-MFI59545.2023.10361452
DatabaseName IEEE Electronic Library (IEL) Conference Proceedings
IEEE Xplore POP ALL
IEEE Xplore All Conference Proceedings
IEEE Electronic Library (IEL)
IEEE Proceedings Order Plans (POP All) 1998-Present
DatabaseTitleList
Database_xml – sequence: 1
  dbid: RIE
  name: IEEE/IET Electronic Library (IEL) (UW System Shared)
  url: https://ieeexplore.ieee.org/
  sourceTypes: Publisher
DeliveryMethod fulltext_linktorsrc
Discipline Engineering
EISBN 9798350382587
EISSN 2767-9357
EndPage 8
ExternalDocumentID 10361452
Genre orig-research
GrantInformation_xml – fundername: Wellcome Trust
  funderid: 10.13039/100010269
GroupedDBID 6IE
6IF
6IK
6IL
6IN
AAJGR
ABLEC
ADZIZ
ALMA_UNASSIGNED_HOLDINGS
BEFXN
BFFAM
BGNUA
BKEBE
BPEOZ
CBEJK
CHZPO
IEGSK
IPLJI
KZ1
M43
OCL
RIE
RIL
RNS
ID FETCH-LOGICAL-i490-a43784cc763aef817c760aaa806ff35e8c15f14c03c2a03a521ad1834f96177c3
IEDL.DBID RIE
IngestDate Wed Aug 27 02:25:30 EDT 2025
IsPeerReviewed false
IsScholarly false
Language English
LinkModel DirectLink
MergedId FETCHMERGED-LOGICAL-i490-a43784cc763aef817c760aaa806ff35e8c15f14c03c2a03a521ad1834f96177c3
PageCount 8
ParticipantIDs ieee_primary_10361452
PublicationCentury 2000
PublicationDate 2023-Nov.-27
PublicationDateYYYYMMDD 2023-11-27
PublicationDate_xml – month: 11
  year: 2023
  text: 2023-Nov.-27
  day: 27
PublicationDecade 2020
PublicationTitle IEEE/SICE/RSJ International Conference on Multisensor Fusion and Integration for Intelligent Systems
PublicationTitleAbbrev SDF-MFI
PublicationYear 2023
Publisher IEEE
Publisher_xml – name: IEEE
SSID ssj0001615956
Score 1.8525916
Snippet Calibrating statistical models using Bayesian inference often requires both accurate and timely estimates of parameters of interest. Particle Markov Chain...
SourceID ieee
SourceType Publisher
StartPage 1
SubjectTerms Bayes methods
Bayesian inference
biomedical applications
distributed memory
Markov processes
Monte Carlo methods
parallel algorithms
Parameter estimation
Particle filters
particle Markov Chain Monte Carlo
Program processors
Recycling
SMC 2
Title An O(log2 N) SMC2 Algorithm on Distributed Memory with an Approx. Optimal L-Kernel
URI https://ieeexplore.ieee.org/document/10361452
hasFullText 1
inHoldings 1
isFullTextHit
isPrint
link http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV3LTgMhFCXauNCNrxrfYeFCF4zMAMOwbFonGvuKbUx3DcNDm7QzprZG_16YVqsLF-4IC0K44R64cM4B4IJjoVgmYmS5iRHNGEaZZhzZyDKsZaJIKfb82OTtdjIYiO6SrF5yYYwx5eczE_hm-ZavCzX3pTK3w4lDE-Yy7jrnfEHWWhVUHDa7w_5SWDTE4rrXSFErvXO9lAXeJjz4GuCXlUqJJOn2P-ewA6orTh7sfqPNLlgz-R7Y-iEnuA8eajnsXHqvV9i-gr1WPYK18VPhrv_PE1jksOFFcr2_ldGw5X_YfkBfhoUyhzUvLf4ewI7LIBM5hk10b6a5GVdBP73p12_R0jMBjajASFLCE6qUyxrS2CTkroWllAmOrSXMJCpkNqQKExVJTKQDb6ndrqZWuKMMV-QAVPIiN4cARsrFysZc08iNGklJQhPrWBCmM8uoOQJVvzrDl4UqxvBrYY7_6D8Bmz4GnscX8VNQmU3n5gxsqLfZ6HV6XsbyEzwgnOQ
linkProvider IEEE
linkToHtml http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV3LTgIxFG0MmqgbXxjfduFCF8VOH9OZJQEJBGYgQgw7UjqtksCMQTD697YDiC5cuGvuoml603ua255zALgROFR8GPrICO0jNuQYDRMukCGG40QGiuZiz08tEcdBvx92lmT1nAujtc4_n-mSG-Zv-Umm5q5VZk84tWjCbcXd5IwRb0HXWrdULDrb6_5SWtTD4X23WkNRrWGjjJecUXhpNcUvM5UcS2p7_1zFPiiuWXmw8403B2BDp4dg94eg4BF4LKewfevcXmF8B7tRhcDy-DmbjmYvE5ilsOpkcp3DlU5g5P7YfkLXiIUyhWUnLv5Rgm1bQyZyDFuoqaepHhdBr_bQq9TR0jUBjViIkWRUBEwpWzekNoEn7AhLKQPsG0O5DpTHjccUpopITKWFb5nYc81MaC8zQtFjUEizVJ8ASJTNlvFFwoidlUhJPe0nfkh5MjSc6VNQdLszeF3oYgxWG3P2R_wabNd7UWvQasTNc7Dj8uFYfURcgMJsOteXYEu9z0Zv06s8r1-Y1qAr
openUrl ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Abook&rft.genre=proceeding&rft.title=IEEE%2FSICE%2FRSJ+International+Conference+on+Multisensor+Fusion+and+Integration+for+Intelligent+Systems&rft.atitle=An+O%28log2+N%29+SMC2+Algorithm+on+Distributed+Memory+with+an+Approx.+Optimal+L-Kernel&rft.au=Rosato%2C+Conor&rft.au=Varsi%2C+Alessandro&rft.au=Murphy%2C+Joshua&rft.au=Maskell%2C+Simon&rft.date=2023-11-27&rft.pub=IEEE&rft.eissn=2767-9357&rft.spage=1&rft.epage=8&rft_id=info:doi/10.1109%2FSDF-MFI59545.2023.10361452&rft.externalDocID=10361452