001031521 001__ 1031521
001031521 005__ 20250203133210.0
001031521 0247_ $$2doi$$a10.1007/s42484-024-00198-5
001031521 0247_ $$2ISSN$$a2524-4906
001031521 0247_ $$2ISSN$$a2524-4914
001031521 0247_ $$2datacite_doi$$a10.34734/FZJ-2024-05716
001031521 0247_ $$2WOS$$aWOS:001312144200001
001031521 037__ $$aFZJ-2024-05716
001031521 041__ $$aEnglish
001031521 082__ $$a050
001031521 1001_ $$0P:(DE-HGF)0$$aWulff, Eric$$b0$$eCorresponding author
001031521 245__ $$aDistributed hybrid quantum-classical performance prediction for hyperparameter optimization
001031521 260__ $$a[Cham]$$bSpringer Nature Switzerland AG$$c2024
001031521 3367_ $$2DRIVER$$aarticle
001031521 3367_ $$2DataCite$$aOutput Types/Journal article
001031521 3367_ $$0PUB:(DE-HGF)16$$2PUB:(DE-HGF)$$aJournal Article$$bjournal$$mjournal$$s1730795673_29934
001031521 3367_ $$2BibTeX$$aARTICLE
001031521 3367_ $$2ORCID$$aJOURNAL_ARTICLE
001031521 3367_ $$00$$2EndNote$$aJournal Article
001031521 520__ $$aHyperparameter optimization (HPO) of neural networks is a computationally expensive procedure, which requires a large number of different model configurations to be trained. To reduce such costs, this work presents a distributed, hybrid workflow, that runs the training of the neural networks on multiple graphics processing units (GPUs) on a classical supercomputer, while predicting the configurations’ performance with quantum-trained support vector regression (QT-SVR) on a quantum annealer (QA). The workflow is shown to run on up to 50 GPUs and a QA at the same time, completely automating the communication between the classical and the quantum systems. The approach is evaluated extensively on several benchmarking datasets from the computer vision (CV), high-energy physics (HEP), and natural language processing (NLP) domains. Empirical results show that resource costs for performing HPO can be reduced by up to 9% when using the hybrid workflow with performance prediction, compared to using a plain HPO algorithm without performance prediction. Additionally, the workflow obtains similar and in some cases even better accuracy of the final hyperparameter configuration, when combining multiple heuristically obtained predictions from the QA, compared to using just a single classically obtained prediction. The results highlight the potential of hybrid quantum-classical machine learning algorithms. The workflow code is made available open-source to foster adoption in the community.
001031521 536__ $$0G:(DE-HGF)POF4-5111$$a5111 - Domain-Specific Simulation & Data Life Cycle Labs (SDLs) and Research Groups (POF4-511)$$cPOF4-511$$fPOF IV$$x0
001031521 536__ $$0G:(EU-Grant)951733$$aRAISE - Research on AI- and Simulation-Based Engineering at Exascale (951733)$$c951733$$fH2020-INFRAEDI-2019-1$$x1
001031521 588__ $$aDataset connected to CrossRef, Journals: juser.fz-juelich.de
001031521 7001_ $$0P:(DE-HGF)0$$aGarcia Amboage, Juan Pablo$$b1
001031521 7001_ $$0P:(DE-Juel1)180916$$aAach, Marcel$$b2
001031521 7001_ $$0P:(DE-HGF)0$$aGislason, Thorsteinn Eli$$b3
001031521 7001_ $$0P:(DE-HGF)0$$aIngolfsson, Thorsteinn Kristinn$$b4
001031521 7001_ $$0P:(DE-HGF)0$$aIngolfsson, Tomas Kristinn$$b5
001031521 7001_ $$0P:(DE-Juel1)191143$$aPasetto, Edoardo$$b6
001031521 7001_ $$0P:(DE-Juel1)191384$$aDelilbasic, Amer$$b7
001031521 7001_ $$0P:(DE-Juel1)132239$$aRiedel, Morris$$b8
001031521 7001_ $$0P:(DE-Juel1)188513$$aSarma, Rakesh$$b9
001031521 7001_ $$0P:(DE-HGF)0$$aGirone, Maria$$b10
001031521 7001_ $$0P:(DE-Juel1)165948$$aLintermann, Andreas$$b11
001031521 773__ $$0PERI:(DE-600)2947496-6$$a10.1007/s42484-024-00198-5$$gVol. 6, no. 2, p. 59$$n2$$p59$$tQuantum machine intelligence$$v6$$x2524-4906$$y2024
001031521 8564_ $$uhttps://juser.fz-juelich.de/record/1031521/files/Wulff%20et%20al._2024_Distributed%20hybrid%20quantum-classical%20performance%20prediction%20for%20hyperparameter%20optimization%282%29.pdf$$yOpenAccess
001031521 8564_ $$uhttps://juser.fz-juelich.de/record/1031521/files/Wulff%20et%20al._2024_Distributed%20hybrid%20quantum-classical%20performance%20prediction%20for%20hyperparameter%20optimization%282%29.gif?subformat=icon$$xicon$$yOpenAccess
001031521 8564_ $$uhttps://juser.fz-juelich.de/record/1031521/files/Wulff%20et%20al._2024_Distributed%20hybrid%20quantum-classical%20performance%20prediction%20for%20hyperparameter%20optimization%282%29.jpg?subformat=icon-1440$$xicon-1440$$yOpenAccess
001031521 8564_ $$uhttps://juser.fz-juelich.de/record/1031521/files/Wulff%20et%20al._2024_Distributed%20hybrid%20quantum-classical%20performance%20prediction%20for%20hyperparameter%20optimization%282%29.jpg?subformat=icon-180$$xicon-180$$yOpenAccess
001031521 8564_ $$uhttps://juser.fz-juelich.de/record/1031521/files/Wulff%20et%20al._2024_Distributed%20hybrid%20quantum-classical%20performance%20prediction%20for%20hyperparameter%20optimization%282%29.jpg?subformat=icon-640$$xicon-640$$yOpenAccess
001031521 909CO $$ooai:juser.fz-juelich.de:1031521$$pdnbdelivery$$pec_fundedresources$$pVDB$$pdriver$$popen_access$$popenaire
001031521 9101_ $$0I:(DE-588b)5008462-8$$6P:(DE-Juel1)180916$$aForschungszentrum Jülich$$b2$$kFZJ
001031521 9101_ $$0I:(DE-588b)5008462-8$$6P:(DE-Juel1)191143$$aForschungszentrum Jülich$$b6$$kFZJ
001031521 9101_ $$0I:(DE-588b)5008462-8$$6P:(DE-Juel1)191384$$aForschungszentrum Jülich$$b7$$kFZJ
001031521 9101_ $$0I:(DE-588b)5008462-8$$6P:(DE-Juel1)132239$$aForschungszentrum Jülich$$b8$$kFZJ
001031521 9101_ $$0I:(DE-588b)5008462-8$$6P:(DE-Juel1)188513$$aForschungszentrum Jülich$$b9$$kFZJ
001031521 9101_ $$0I:(DE-588b)5008462-8$$6P:(DE-Juel1)165948$$aForschungszentrum Jülich$$b11$$kFZJ
001031521 9131_ $$0G:(DE-HGF)POF4-511$$1G:(DE-HGF)POF4-510$$2G:(DE-HGF)POF4-500$$3G:(DE-HGF)POF4$$4G:(DE-HGF)POF$$9G:(DE-HGF)POF4-5111$$aDE-HGF$$bKey Technologies$$lEngineering Digital Futures – Supercomputing, Data Management and Information Security for Knowledge and Action$$vEnabling Computational- & Data-Intensive Science and Engineering$$x0
001031521 9141_ $$y2024
001031521 915__ $$0LIC:(DE-HGF)CCBY4$$2HGFVOC$$aCreative Commons Attribution CC BY 4.0
001031521 915__ $$0StatID:(DE-HGF)3002$$2StatID$$aDEAL Springer$$d2023-10-27$$wger
001031521 915__ $$0StatID:(DE-HGF)0510$$2StatID$$aOpenAccess
001031521 915__ $$0StatID:(DE-HGF)0100$$2StatID$$aJCR$$bQUANT MACH INTELL : 2022$$d2024-12-05
001031521 915__ $$0StatID:(DE-HGF)0200$$2StatID$$aDBCoverage$$bSCOPUS$$d2024-12-05
001031521 915__ $$0StatID:(DE-HGF)0300$$2StatID$$aDBCoverage$$bMedline$$d2024-12-05
001031521 915__ $$0StatID:(DE-HGF)0199$$2StatID$$aDBCoverage$$bClarivate Analytics Master Journal List$$d2024-12-05
001031521 915__ $$0StatID:(DE-HGF)0112$$2StatID$$aWoS$$bEmerging Sources Citation Index$$d2024-12-05
001031521 915__ $$0StatID:(DE-HGF)0150$$2StatID$$aDBCoverage$$bWeb of Science Core Collection$$d2024-12-05
001031521 915__ $$0StatID:(DE-HGF)9900$$2StatID$$aIF < 5$$d2024-12-05
001031521 920__ $$lyes
001031521 9201_ $$0I:(DE-Juel1)JSC-20090406$$kJSC$$lJülich Supercomputing Center$$x0
001031521 980__ $$ajournal
001031521 980__ $$aVDB
001031521 980__ $$aUNRESTRICTED
001031521 980__ $$aI:(DE-Juel1)JSC-20090406
001031521 9801_ $$aFullTexts