Temporal Parallelization of Inference in Hidden Markov Models

This paper presents algorithms for the parallelization of inference in hidden Markov models (HMMs). In particular, we propose a parallel forward-backward type of filtering and smoothing algorithm as well as a parallel Viterbi-type maximum-a-posteriori (MAP) algorithm. We define associative elements...

Celý popis

Uložené v:

Podrobná bibliografia
Vydané v:	IEEE transactions on signal processing Ročník 69; s. 4875 - 4887
Hlavní autori:	Hassan, Sakira, Sarkka, Simo, Garcia-Fernandez, Angel
Médium:	Journal Article
Jazyk:	English
Vydavateľské údaje:	New York IEEE 2021 The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Predmet:	Algorithms Graphics processing units Hidden Markov models Inference Inference algorithms Markov chains Parallel forward-backward algorithm parallel max-product algorithm Parallel processing parallel sum-product algorithm parallel Viterbi algorithm Polytopes Signal processing algorithms Sum product algorithm Task analysis Viterbi algorithm
ISSN:	1053-587X, 1941-0476
On-line prístup:	Získať plný text
Tagy:	Pridať tag Žiadne tagy, Buďte prvý, kto otaguje tento záznam!

Abstract	This paper presents algorithms for the parallelization of inference in hidden Markov models (HMMs). In particular, we propose a parallel forward-backward type of filtering and smoothing algorithm as well as a parallel Viterbi-type maximum-a-posteriori (MAP) algorithm. We define associative elements and operators to pose these inference problems as all-prefix-sums computations and parallelize them using the parallel-scan algorithm. The advantage of the proposed algorithms is that they are computationally efficient in HMM inference problems with long time horizons. We empirically compare the performance of the proposed methods to classical methods on a highly parallel graphics processing unit (GPU).
AbstractList	This paper presents algorithms for the parallelization of inference in hidden Markov models (HMMs). In particular, we propose a parallel forward-backward type of filtering and smoothing algorithm as well as a parallel Viterbi-type maximum-a-posteriori (MAP) algorithm. We define associative elements and operators to pose these inference problems as all-prefix-sums computations and parallelize them using the parallel-scan algorithm. The advantage of the proposed algorithms is that they are computationally efficient in HMM inference problems with long time horizons. We empirically compare the performance of the proposed methods to classical methods on a highly parallel graphics processing unit (GPU).
Author	Hassan, Sakira Garcia-Fernandez, Angel Sarkka, Simo
Author_xml	– sequence: 1 givenname: Sakira orcidid: 0000-0001-6838-474X surname: Hassan fullname: Hassan, Sakira email: syeda.s.hassan@aalto.fi organization: Department of Electrical Engineering and Automation, Aalto University, Espoo, Finland – sequence: 2 givenname: Simo orcidid: 0000-0002-7031-9354 surname: Sarkka fullname: Sarkka, Simo email: simo.sarkka@aalto.fi organization: Department of Electrical Engineering and Automation, Aalto University, Espoo, Finland – sequence: 3 givenname: Angel orcidid: 0000-0002-6471-8455 surname: Garcia-Fernandez fullname: Garcia-Fernandez, Angel email: angel.f.garcia.fernandez@gmail.com organization: Department of Electrical Engineering and Electronics, University of Liverpool, Liverpool, U.K
BookMark	eNp9kM1LAzEQxYNUsK3eBS8LnrdOPnaze_AgRW2hxYIVvIV0M4HUbVKzW0H_ere2ePDgZWYO7_dm5g1IzwePhFxSGFEK5c3yeTFiwOiIU-CcFyekT0tBUxAy73UzZDzNCvl6RgZNswagQpR5n9wucbMNUdfJQne1xtp96dYFnwSbTL3FiL7CxPlk4oxBn8x1fAsfyTwYrJtzcmp13eDFsQ_Jy8P9cjxJZ0-P0_HdLK26U9pUrgAzwzS3VS4LZg12VhKl4CAApSygNBSkrSSYQq-MzQtmDEchqCnKFedDcn3w3cbwvsOmVeuwi75bqViWF1JKlu1V-UFVxdA0Ea2qXPvzTBu1qxUFtY9KdVGpfVTqGFUHwh9wG91Gx8__kKsD4hDxV15mlPFS8m_GTXWZ
CODEN	ITPRED
CitedBy_id	crossref_primary_10_1109_LSP_2024_3519258 crossref_primary_10_1109_TAC_2024_3518309 crossref_primary_10_1109_TAI_2024_3358261 crossref_primary_10_1137_23M156121X crossref_primary_10_1109_ACCESS_2025_3562686 crossref_primary_10_1109_TAC_2022_3147017 crossref_primary_10_21105_joss_07069 crossref_primary_10_1016_j_csda_2024_107955
Cites_doi	10.1109/TSP.2017.2713759 10.1109/MICRO.2012.48 10.1007/s00453-007-9128-0 10.7551/mitpress/3348.001.0001 10.1109/PDMC-HiBi.2010.24 10.1109/TSP.2013.2277833 10.1109/TAC.1966.1098348 10.1109/TSP.2019.2941064 10.1111/j.2517-6161.1977.tb01600.x 10.1109/MASSP.1986.1165342 10.1109/ICISE.2009.265 10.1017/CBO9781139344203 10.1109/TSP.2018.2869119 10.1007/s11831-020-09422-4 10.1145/322217.322232 10.1007/BF01531015 10.1162/089976600300015880 10.1109/TAC.2020.2976316 10.1109/TSP.2011.2167616 10.1109/5.18626 10.1117/12.707333 10.1109/IPDPS.2011.181 10.1145/2692916.2555264 10.1145/3079856.3080246 10.1080/00401706.1991.10484833 10.1006/jmbi.2000.4315 10.1109/JPROC.2008.917757 10.1109/TSP.2015.2461518 10.1109/TIT.1967.1054010 10.1214/aoms/1177697196 10.1109/12.42122 10.1109/ISBI.2008.4541126 10.1109/TSP.2013.2256905
ContentType	Journal Article
Copyright	Copyright The Institute of Electrical and Electronics Engineers, Inc. (IEEE) 2021
Copyright_xml	– notice: Copyright The Institute of Electrical and Electronics Engineers, Inc. (IEEE) 2021
DBID	97E ESBDL RIA RIE AAYXX CITATION 7SC 7SP 8FD JQ2 L7M L~C L~D
DOI	10.1109/TSP.2021.3103338
DatabaseName	IEEE All-Society Periodicals Package (ASPP) 2005–Present IEEE Xplore Open Access Journals (WRLC) IEEE All-Society Periodicals Package (ASPP) 1998–Present IEEE Xplore CrossRef Computer and Information Systems Abstracts Electronics & Communications Abstracts Technology Research Database ProQuest Computer Science Collection Advanced Technologies Database with Aerospace Computer and Information Systems Abstracts Academic Computer and Information Systems Abstracts Professional
DatabaseTitle	CrossRef Technology Research Database Computer and Information Systems Abstracts – Academic Electronics & Communications Abstracts ProQuest Computer Science Collection Computer and Information Systems Abstracts Advanced Technologies Database with Aerospace Computer and Information Systems Abstracts Professional
DatabaseTitleList	Technology Research Database
Database_xml	– sequence: 1 dbid: RIE name: IEEE Xplore url: https://ieeexplore.ieee.org/ sourceTypes: Publisher
DeliveryMethod	fulltext_linktorsrc
Discipline	Engineering
EISSN	1941-0476
EndPage	4887
ExternalDocumentID	10_1109_TSP_2021_3103338 9512397
Genre	orig-research
GrantInformation_xml	– fundername: Academy of Finland funderid: 10.13039/501100002341
GroupedDBID	-~X .DC 0R~ 29I 3EH 4.4 53G 5GY 5VS 6IK 85S 97E AAJGR AARMG AASAJ AAWTH ABAZT ABFSI ABQJQ ABVLG ACGFO ACIWK ACKIV ACNCT AENEX AETIX AGQYO AGSQL AHBIQ AI. AIBXA AJQPL AKJIK AKQYR ALLEH ALMA_UNASSIGNED_HOLDINGS ASUFR ATWAV BEFXN BFFAM BGNUA BKEBE BPEOZ CS3 E.L EBS EJD ESBDL F5P HZ~ H~9 ICLAB IFIPE IFJZH IPLJI JAVBF LAI MS~ O9- OCL P2P RIA RIE RNS TAE TN5 VH1 AAYXX CITATION 7SC 7SP 8FD JQ2 L7M L~C L~D
ID	FETCH-LOGICAL-c333t-7b0e5d2a3fc6782fdedde7e743040e77809d107fc70d8abdf682dd3e441d89b33
IEDL.DBID	RIE
ISICitedReferencesCount	10
ISICitedReferencesURI	http://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=Summon&SrcAuth=ProQuest&DestLinkType=CitingArticles&DestApp=WOS_CPL&KeyUT=000692231000008&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
ISSN	1053-587X
IngestDate	Mon Jun 30 10:20:10 EDT 2025 Tue Nov 18 22:31:04 EST 2025 Sat Nov 29 04:10:54 EST 2025 Wed Aug 27 02:27:33 EDT 2025
IsDoiOpenAccess	true
IsOpenAccess	true
IsPeerReviewed	true
IsScholarly	true
Language	English
License	https://creativecommons.org/licenses/by/4.0/legalcode
LinkModel	DirectLink
MergedId	FETCHMERGED-LOGICAL-c333t-7b0e5d2a3fc6782fdedde7e743040e77809d107fc70d8abdf682dd3e441d89b33
Notes	ObjectType-Article-1 SourceType-Scholarly Journals-1 ObjectType-Feature-2 content type line 14
ORCID	0000-0001-6838-474X 0000-0002-6471-8455 0000-0002-7031-9354
OpenAccessLink	https://ieeexplore.ieee.org/document/9512397
PQID	2568777253
PQPubID	85478
PageCount	13
ParticipantIDs	proquest_journals_2568777253 crossref_citationtrail_10_1109_TSP_2021_3103338 ieee_primary_9512397 crossref_primary_10_1109_TSP_2021_3103338
PublicationCentury	2000
PublicationDate	20210000 2021-00-00 20210101
PublicationDateYYYYMMDD	2021-01-01
PublicationDate_xml	– year: 2021 text: 20210000
PublicationDecade	2020
PublicationPlace	New York
PublicationPlace_xml	– name: New York
PublicationTitle	IEEE transactions on signal processing
PublicationTitleAbbrev	TSP
PublicationYear	2021
Publisher	IEEE The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Publisher_xml	– name: IEEE – name: The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
References	ref35 ref13 ref34 ref12 ref15 ref36 ref14 ref30 ref11 ref32 ref10 blelloch (ref25) 1990 ref1 ref38 ref16 ref19 cappé (ref2) 2006 cowell (ref40) 2006 pearl (ref37) 1988 hymel (ref29) 2011 abadi (ref46) 2015 koller (ref33) 2009 bishop (ref39) 2006 ref24 chatterjee (ref17) 2011 ref45 ref23 li (ref26) 2009; 4 ref20 ref42 ref41 ref22 ref44 du (ref18) 2010 ref21 ref43 ref28 yaghoobi (ref31) 2021 ref8 ref7 ref9 ref4 ref3 liu (ref27) 2009 ref6 ref5
References_xml	– year: 1990 ident: ref25 article-title: Prefix sums and their applications – ident: ref34 doi: 10.1109/TSP.2017.2713759 – ident: ref21 doi: 10.1109/MICRO.2012.48 – ident: ref14 doi: 10.1007/s00453-007-9128-0 – ident: ref41 doi: 10.7551/mitpress/3348.001.0001 – ident: ref15 doi: 10.1109/PDMC-HiBi.2010.24 – ident: ref3 doi: 10.1109/TSP.2013.2277833 – year: 2009 ident: ref33 publication-title: Probabilistic Graphical Models Principles and Techniques – ident: ref43 doi: 10.1109/TAC.1966.1098348 – year: 2006 ident: ref2 publication-title: Inference in Hidden Markov Models – ident: ref12 doi: 10.1109/TSP.2019.2941064 – year: 2015 ident: ref46 article-title: TensorFlow: Large-scale machine learning on heterogeneous systems – ident: ref45 doi: 10.1111/j.2517-6161.1977.tb01600.x – ident: ref36 doi: 10.1109/MASSP.1986.1165342 – start-page: 1 year: 2010 ident: ref18 article-title: A tile-based parallel Viterbi algorithm for biological sequence alignment on GPU with CUDA publication-title: Proc IEEE Int Symp Parallel Distrib Process Workshops PhD Forum – ident: ref28 doi: 10.1109/ICISE.2009.265 – ident: ref32 doi: 10.1017/CBO9781139344203 – ident: ref10 doi: 10.1109/TSP.2018.2869119 – ident: ref5 doi: 10.1007/s11831-020-09422-4 – ident: ref42 doi: 10.1145/322217.322232 – ident: ref38 doi: 10.1007/BF01531015 – ident: ref35 doi: 10.1162/089976600300015880 – start-page: 96 year: 2011 ident: ref17 article-title: A temporally abstracted Viterbi algorithm publication-title: Proc Conf Uncertainty of Artificial Intelligence – ident: ref30 doi: 10.1109/TAC.2020.2976316 – ident: ref11 doi: 10.1109/TSP.2011.2167616 – year: 2009 ident: ref27 article-title: cuHMM: A CUDA implementation of hidden markov model training and classification – year: 2011 ident: ref29 article-title: Massively parallel hidden Markov models for wireless applications – ident: ref1 doi: 10.1109/5.18626 – ident: ref7 doi: 10.1117/12.707333 – ident: ref16 doi: 10.1109/IPDPS.2011.181 – ident: ref19 doi: 10.1145/2692916.2555264 – ident: ref22 doi: 10.1145/3079856.3080246 – ident: ref6 doi: 10.1080/00401706.1991.10484833 – ident: ref8 doi: 10.1006/jmbi.2000.4315 – year: 2006 ident: ref39 publication-title: Pattern Recognition and Machine Learning – volume: 4 start-page: 426 year: 2009 ident: ref26 article-title: The fast evaluation of hidden Markov models on GPU publication-title: Proc IEEE Int Conf Intell Comput Intell Syst – start-page: 5350 year: 2021 ident: ref31 article-title: Parallel iterated extended and sigma-point Kalman smoothers publication-title: Proc ICASSP IEEE Int Conf Acoustics Speech Signal Process – ident: ref20 doi: 10.1109/JPROC.2008.917757 – year: 2006 ident: ref40 publication-title: Probabilistic Networks and Expert Systems Exact Computational Methods for Bayesian Networks – ident: ref9 doi: 10.1109/TSP.2015.2461518 – ident: ref13 doi: 10.1109/TIT.1967.1054010 – ident: ref44 doi: 10.1214/aoms/1177697196 – ident: ref24 doi: 10.1109/12.42122 – year: 1988 ident: ref37 publication-title: Probabilistic Reasoning in Intelligent Systems Networks of Plausible Inference – ident: ref23 doi: 10.1109/ISBI.2008.4541126 – ident: ref4 doi: 10.1109/TSP.2013.2256905
SSID	ssj0014496
Score	2.4479463
Snippet	This paper presents algorithms for the parallelization of inference in hidden Markov models (HMMs). In particular, we propose a parallel forward-backward type...
SourceID	proquest crossref ieee
SourceType	Aggregation Database Enrichment Source Index Database Publisher
StartPage	4875
SubjectTerms	Algorithms Graphics processing units Hidden Markov models Inference Inference algorithms Markov chains Parallel forward-backward algorithm parallel max-product algorithm Parallel processing parallel sum-product algorithm parallel Viterbi algorithm Polytopes Signal processing algorithms Sum product algorithm Task analysis Viterbi algorithm
Title	Temporal Parallelization of Inference in Hidden Markov Models
URI	https://ieeexplore.ieee.org/document/9512397 https://www.proquest.com/docview/2568777253
Volume	69
WOSCitedRecordID	wos000692231000008&url=https%3A%2F%2Fcvtisr.summon.serialssolutions.com%2F%23%21%2Fsearch%3Fho%3Df%26include.ft.matches%3Dt%26l%3Dnull%26q%3D
hasFullText	1
inHoldings	1
isFullTextHit
isPrint
journalDatabaseRights	– providerCode: PRVIEE databaseName: IEEE Xplore customDbUrl: eissn: 1941-0476 dateEnd: 99991231 omitProxy: false ssIdentifier: ssj0014496 issn: 1053-587X databaseCode: RIE dateStart: 19910101 isFulltext: true titleUrlDefault: https://ieeexplore.ieee.org/ providerName: IEEE
link	http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV07a8MwED6S0KEd-kpL0xcauhSqRJHsSB46lNKQLiHQFLIZW5IhYOyS1-_vyS8CLYVuHiRjf7LuvtP57gN4GHoJSzwZUI2GgHpDpakShtPY-sxDh6mZHxViE3I6VYtFMGvBU1MLY60tfj6zfXdZ5PJNrrfuqGyAbICj_2xDW0pZ1mo1GQPPK7S4kC4I6iu5qFOSLBjMP2YYCPJh32lqCVeJsueCCk2VH4a48C7jk_891ykcVyySvJTLfgYtm53D0V5vwS48z8umUymZRSsnmJJWFZckT8h7XeZHlhmZuC4iGXFVO_mOOHG0dH0Bn-O3-euEVloJVOMrbaiMmfUNj0Si0f3wxFicKi3yA9ylVkrFAoORXqIlMyqKTTJS3BhhkQ0ZFcRCXEInyzN7BQQjjpgHJhih_fES5UUMOcnIYhzILTdS9GBQwxfqqpG407NIwyKgYEGIgIcO8LACvAePzYyvsonGH2O7DuBmXIVtD27rFQqrXbYOka65dobcF9e_z7qBQ3fv8sjkFjqb1dbewYHebZbr1X3xAX0DsEnBWA
linkProvider	IEEE
linkToHtml	http://cvtisr.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwlV3PS8MwFH7MKagHf01xOjUHL4LdsiRd0oMHEceGcwycsFtpkxQGo5P9-vt9absxUARvPSS0_dK89728vvcB3DdFQhMhA0-jIfBEU2lPccO82PpUoMPU1I8ysQnZ76vRKBiU4HFTC2OtzX4-s3V3meXyzVQv3VFZA9kAQ_-5A7u-EKyZV2ttcgZCZGpcSBi45ys5WicladAYfgwwFGTNulPV4q4WZcsJZaoqP0xx5l_ax_97shM4Kngkec4X_hRKNj2Dw63ughV4GuZtpyZkEM2cZMqkqLkk04R014V-ZJySjusjkhJXtzNdESePNpmfw2f7dfjS8Qq1BE_jKy08GVPrGxbxRKMDYomxOFVaZAi4T62UigYGY71ES2pUFJukpZgx3CIfMiqIOb-AcjpN7SUQjDliFpighRZIJEpEFFlJy2IkyCwzklehsYYv1EUrcadoMQmzkIIGIQIeOsDDAvAqPGxmfOVtNP4YW3EAb8YV2Fahtl6hsNhn8xAJm2toyHx-9fusO9jvDN97Ya_bf7uGA3ef_AClBuXFbGlvYE-vFuP57Db7mL4B7LLEnw
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%3Ajournal&rft.genre=article&rft.atitle=Temporal+Parallelization+of+Inference+in+Hidden+Markov+Models&rft.jtitle=IEEE+transactions+on+signal+processing&rft.au=Hassan%2C+Sakira&rft.au=Sarkka%2C+Simo&rft.au=Garcia-Fernandez%2C+Angel&rft.date=2021&rft.pub=IEEE&rft.issn=1053-587X&rft.volume=69&rft.spage=4875&rft.epage=4887&rft_id=info:doi/10.1109%2FTSP.2021.3103338&rft.externalDocID=9512397
thumbnail_l	http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/lc.gif&issn=1053-587X&client=summon
thumbnail_m	http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/mc.gif&issn=1053-587X&client=summon
thumbnail_s	http://covers-cdn.summon.serialssolutions.com/index.aspx?isbn=/sc.gif&issn=1053-587X&client=summon