{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,6,26]],"date-time":"2026-06-26T22:57:48Z","timestamp":1782514668224,"version":"3.54.5"},"reference-count":51,"publisher":"IOP Publishing","issue":"1","license":[{"start":{"date-parts":[[2021,12,27]],"date-time":"2021-12-27T00:00:00Z","timestamp":1640563200000},"content-version":"vor","delay-in-days":0,"URL":"http:\/\/creativecommons.org\/licenses\/by\/4.0"},{"start":{"date-parts":[[2021,12,27]],"date-time":"2021-12-27T00:00:00Z","timestamp":1640563200000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/iopscience.iop.org\/info\/page\/text-and-data-mining"}],"funder":[{"name":"European Research Council"},{"DOI":"10.13039\/501100000266","name":"EPSRC","doi-asserted-by":"crossref","award":["EP\/P015794\/1"],"award-info":[{"award-number":["EP\/P015794\/1"]}],"id":[{"id":"10.13039\/501100000266","id-type":"DOI","asserted-by":"crossref"}]},{"DOI":"10.13039\/501100000780","name":"European Union","doi-asserted-by":"crossref","award":["805162"],"award-info":[{"award-number":["805162"]}],"id":[{"id":"10.13039\/501100000780","id-type":"DOI","asserted-by":"crossref"}]},{"DOI":"10.13039\/501100000270","name":"Natural Environment Research Council","doi-asserted-by":"crossref","award":["NE\/P013406\/1"],"award-info":[{"award-number":["NE\/P013406\/1"]}],"id":[{"id":"10.13039\/501100000270","id-type":"DOI","asserted-by":"crossref"}]},{"name":"AWE plc","award":["EP\/M022331\/1"],"award-info":[{"award-number":["EP\/M022331\/1"]}]}],"content-domain":{"domain":["iopscience.iop.org"],"crossmark-restriction":false},"short-container-title":["Mach. Learn.: Sci. Technol."],"published-print":{"date-parts":[[2022,3,1]]},"abstract":"<jats:title>Abstract<\/jats:title>\n               <jats:p>Computer simulations are invaluable tools for scientific discovery. However, accurate simulations are often slow to execute, which limits their applicability to extensive parameter exploration, large-scale data analysis, and uncertainty quantification. A promising route to accelerate simulations by building fast emulators with machine learning requires large training datasets, which can be prohibitively expensive to obtain with slow simulations. Here we present a method based on neural architecture search to build accurate emulators even with a limited number of training data. The method successfully emulates simulations in 10 scientific cases including astrophysics, climate science, biogeochemistry, high energy density physics, fusion energy, and seismology, using the same super-architecture, algorithm, and hyperparameters. Our approach also inherently provides emulator uncertainty estimation, adding further confidence in their use. We anticipate this work will accelerate research involving expensive simulations, allow more extensive parameters exploration, and enable new, previously unfeasible computational discovery.<\/jats:p>","DOI":"10.1088\/2632-2153\/ac3ffa","type":"journal-article","created":{"date-parts":[[2021,12,7]],"date-time":"2021-12-07T14:13:52Z","timestamp":1638886432000},"page":"015013","update-policy":"https:\/\/doi.org\/10.1088\/crossmark-policy","source":"Crossref","is-referenced-by-count":79,"title":["Building high accuracy emulators for scientific simulations with deep neural architecture search"],"prefix":"10.1088","volume":"3","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-8748-1737","authenticated-orcid":false,"given":"M F","family":"Kasim","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"D","family":"Watson-Parris","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"L","family":"Deaconu","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"S","family":"Oliver","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"P","family":"Hatfield","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"D H","family":"Froula","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"G","family":"Gregori","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"M","family":"Jarvis","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"S","family":"Khatiwala","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"J","family":"Korenaga","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"J","family":"Topp-Mugglestone","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0001-6419-6848","authenticated-orcid":false,"given":"E","family":"Viezzer","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"S M","family":"Vinko","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"266","published-online":{"date-parts":[[2021,12,27]]},"reference":[{"key":"mlstac3ffabib1","doi-asserted-by":"publisher","first-page":"909","DOI":"10.1038\/nmat1752","article-title":"Computational high-throughput screening of electrocatalytic materials for hydrogen evolution","volume":"5","author":"Greeley","year":"2006","journal-title":"Nat. Mater."},{"key":"mlstac3ffabib2","doi-asserted-by":"publisher","DOI":"10.1103\/PhysRevLett.102.115001","article-title":"X-ray Thomson-scattering measurements of density and temperature in shock-compressed beryllium","volume":"102","author":"Lee","year":"2009","journal-title":"Phys. Rev. Lett."},{"key":"mlstac3ffabib3","doi-asserted-by":"publisher","DOI":"10.1103\/PhysRevLett.121.025002","article-title":"Beam-ion acceleration during edge localized modes in the ASDEX Upgrade tokamak","volume":"121","author":"Galdon-Quiroga","year":"2018","journal-title":"Phys. Rev. Lett."},{"key":"mlstac3ffabib4","doi-asserted-by":"publisher","DOI":"10.1063\/1.4977912","article-title":"Zonal flow generation in inertial confinement fusion implosions","volume":"24","author":"Peterson","year":"2017","journal-title":"Phys. Plasmas"},{"key":"mlstac3ffabib5","doi-asserted-by":"publisher","first-page":"35","DOI":"10.1088\/0004-637X\/810\/1\/35","article-title":"Cosmic emulation: fast predictions for the galaxy power spectrum","volume":"810","author":"Kwan","year":"2015","journal-title":"Astrophys. J."},{"key":"mlstac3ffabib6","doi-asserted-by":"publisher","first-page":"872","DOI":"10.1038\/s41467-017-00839-3","article-title":"Bypassing the Kohn-Sham equations with machine learning","volume":"8","author":"Brockherde","year":"2017","journal-title":"Nat. Commun."},{"key":"mlstac3ffabib7","doi-asserted-by":"publisher","DOI":"10.1103\/PhysRevLett.108.058301","article-title":"Fast and accurate modeling of molecular atomization energies with machine learning","volume":"108","author":"Rupp","year":"2012","journal-title":"Phys. Rev. Lett."},{"key":"mlstac3ffabib8","first-page":"pp 9446","article-title":"Deep image prior","author":"Ulyanov","year":"2018"},{"key":"mlstac3ffabib9","article-title":"Efficient neural architecture search via parameter sharing","author":"Pham","year":"2018"},{"key":"mlstac3ffabib10","article-title":"Proxylessnas: direct neural architecture search on target task and hardware","author":"Cai","year":"2018"},{"key":"mlstac3ffabib11","first-page":"pp 770","article-title":"Deep residual learning for image recognition","author":"He","year":"2016"},{"key":"mlstac3ffabib12","first-page":"pp 492","author":"Huber","year":"1992"},{"key":"mlstac3ffabib13","doi-asserted-by":"publisher","first-page":"229","DOI":"10.1007\/BF00992696","article-title":"Simple statistical gradient-following algorithms for connectionist reinforcement learning","volume":"8","author":"Williams","year":"1992","journal-title":"Mach. Learn."},{"key":"mlstac3ffabib14","article-title":"The CMA evolution strategy: a tutorial","author":"Hansen","year":"2016"},{"key":"mlstac3ffabib15","doi-asserted-by":"publisher","DOI":"10.1103\/PhysRevE.67.026412","article-title":"Theoretical model of x-ray scattering as a dense matter probe","volume":"67","author":"Gregori","year":"2003","journal-title":"Phys. Rev. E"},{"key":"mlstac3ffabib16","doi-asserted-by":"publisher","first-page":"591","DOI":"10.1038\/s41467-018-02953-2","article-title":"Laboratory evidence of dynamo amplification of magnetic fields in a turbulent plasma","volume":"9","author":"Tzeferacos","year":"2018","journal-title":"Nat. Commun."},{"key":"mlstac3ffabib17","doi-asserted-by":"publisher","DOI":"10.1103\/PhysRevLett.111.045001","article-title":"Hot-spot mix in ignition-scale inertial confinement fusion targets","volume":"111","author":"Regan","year":"2013","journal-title":"Phys. Rev. Lett."},{"key":"mlstac3ffabib18","doi-asserted-by":"publisher","DOI":"10.1063\/1.5000774","article-title":"Simultaneous diagnosis of radial profiles and mix in NIF ignition-scale implosions via x-ray spectroscopy","volume":"24","author":"Ciricosta","year":"2017","journal-title":"Phys. Plasmas"},{"key":"mlstac3ffabib19","doi-asserted-by":"publisher","first-page":"2618","DOI":"10.1093\/mnras\/stw769","article-title":"The galaxy\u2013halo connection in the VIDEO survey at 0.5<z<1.7","volume":"459","author":"Hatfield","year":"2016","journal-title":"Mon. Not. R. Astron. Soc."},{"key":"mlstac3ffabib20","doi-asserted-by":"publisher","DOI":"10.1029\/2012JB009248","article-title":"Seismic tomography of Shatsky Rise by adaptive importance sampling","volume":"117","author":"Korenaga","year":"2012","journal-title":"J. Geophys. Res. Solid Earth"},{"key":"mlstac3ffabib21","doi-asserted-by":"publisher","first-page":"1643","DOI":"10.5194\/gmd-12-1643-2019","article-title":"The global aerosol-climate model ECHAM6.3-HAM2.3\u2014part 1: aerosol evaluation","volume":"12","author":"Tegen","year":"2019","journal-title":"Geosci. Model Dev."},{"key":"mlstac3ffabib22","doi-asserted-by":"publisher","first-page":"GB3001","DOI":"10.1029\/2007GB002923","article-title":"A computational framework for simulation of biogeochemical tracers in the ocean","volume":"21","author":"Khatiwala","year":"2007","journal-title":"Glob. Biogeochem. Cycles"},{"key":"mlstac3ffabib23","article-title":"Cycle consistent surrogate for inertial confinement fusion","author":"Anirudh","year":"2019"},{"key":"mlstac3ffabib24","article-title":"CMA-ES for hyperparameter optimization of deep neural networks","author":"Loshchilov","year":"2016"},{"key":"mlstac3ffabib25","first-page":"2825","article-title":"Scikit-learn: machine learning in python","volume":"12","author":"Pedregosa","year":"2011","journal-title":"J. Mach. Learn. Res."},{"key":"mlstac3ffabib26","doi-asserted-by":"publisher","first-page":"9741","DOI":"10.1073\/pnas.1916634117","article-title":"Improved surrogates in inertial confinement fusion with manifold and cycle consistencies","volume":"117","author":"Anirudh","year":"2020","journal-title":"Proc. Natl Acad. Sci."},{"key":"mlstac3ffabib27","doi-asserted-by":"publisher","DOI":"10.1063\/1.5125979","article-title":"Inverse problem instabilities in large-scale modeling of matter in extreme conditions","volume":"26","author":"Kasim","year":"2019","journal-title":"Phys. Plasmas"},{"key":"mlstac3ffabib28","first-page":"pp 3381","article-title":"Natural evolution strategies","author":"Wierstra","year":"2008"},{"key":"mlstac3ffabib29","doi-asserted-by":"publisher","first-page":"65","DOI":"10.2140\/camcos.2010.5.65","article-title":"Ensemble samplers with affine invariance","volume":"5","author":"Goodman","year":"2010","journal-title":"Commun. Appl. Math. Comput. Sci."},{"key":"mlstac3ffabib30","first-page":"1929","article-title":"Dropout: a simple way to prevent neural networks from overfitting","volume":"15","author":"Srivastava","year":"2014","journal-title":"J. Mach. Learn. Res."},{"key":"mlstac3ffabib31","first-page":"pp 1050","article-title":"Dropout as a Bayesian approximation: representing model uncertainty in deep learning","author":"Gal","year":"2016"},{"key":"mlstac3ffabib32","first-page":"pp 3581","article-title":"Concrete dropout","author":"Gal","year":"2017"},{"key":"mlstac3ffabib33","article-title":"UMAP: uniform manifold approximation and projection for dimension reduction","author":"McInnes","year":"2018"},{"key":"mlstac3ffabib34","first-page":"pp 4763","article-title":"The convergence rate of neural networks for learned functions of different frequencies","author":"Ronen","year":"2019"},{"key":"mlstac3ffabib35","article-title":"Efficient parameter sampling for neural network construction","author":"Chowdhury","year":"2019"},{"key":"mlstac3ffabib36","doi-asserted-by":"publisher","first-page":"526","DOI":"10.1038\/s41586-019-1116-4","article-title":"Predicting disruptive instabilities in controlled fusion plasmas through deep learning","volume":"568","author":"Kates-Harbeck","year":"2019","journal-title":"Nature"},{"key":"mlstac3ffabib37","doi-asserted-by":"publisher","DOI":"10.1103\/PhysRevAccelBeams.21.112802","article-title":"Machine learning-based longitudinal phase space prediction of particle accelerators","volume":"21","author":"Emma","year":"2018","journal-title":"Phys. Rev. Accel. Beams"},{"key":"mlstac3ffabib38","doi-asserted-by":"publisher","first-page":"180","DOI":"10.1109\/TEVC.2013.2248012","article-title":"A Gaussian process surrogate model assisted evolutionary algorithm for medium scale expensive optimization problems","volume":"18","author":"Liu","year":"2013","journal-title":"IEEE Trans. Evol. Comput."},{"key":"mlstac3ffabib39","doi-asserted-by":"publisher","first-page":"148","DOI":"10.1109\/JPROC.2015.2494218","article-title":"Taking the human out of the loop: a review of bayesian optimization","volume":"104","author":"Shahriari","year":"2015","journal-title":"Proc. IEEE"},{"key":"mlstac3ffabib40","doi-asserted-by":"publisher","first-page":"69","DOI":"10.1126\/science.1161466","article-title":"Ultrafast x-ray Thomson scattering of shock-compressed matter","volume":"322","author":"Kritcher","year":"2008","journal-title":"Science"},{"key":"mlstac3ffabib41","doi-asserted-by":"publisher","first-page":"105","DOI":"10.1088\/0741-3335\/38\/2\/001","article-title":"Edge localized modes (ELMs)","volume":"38","author":"Zohm","year":"1996","journal-title":"Plasma Phys. Control. Fusion"},{"key":"mlstac3ffabib42","doi-asserted-by":"publisher","DOI":"10.1088\/1361-6587\/aa7ad0","article-title":"Pedestal and Er profile evolution during an edge localized mode cycle at ASDEX Upgrade","volume":"59","author":"Cavedon","year":"2017","journal-title":"Plasma Phys. Control. Fusion"},{"key":"mlstac3ffabib43","doi-asserted-by":"publisher","DOI":"10.1088\/1741-4326\/aaa22f","article-title":"Ion heat transport dynamics during edge localized mode cycles at ASDEX Upgrade","volume":"58","author":"Viezzer","year":"2018","journal-title":"Nucl. Fusion"},{"key":"mlstac3ffabib44","doi-asserted-by":"publisher","DOI":"10.1088\/0741-3335\/55\/12\/124028","article-title":"Novel free-boundary equilibrium and transport solver with theory-based models and its validation against ASDEX Upgrade current ramp scenarios","volume":"55","author":"Fable","year":"2013","journal-title":"Plasma Phys. Control. Fusion"},{"key":"mlstac3ffabib45","doi-asserted-by":"publisher","DOI":"10.1088\/0029-5515\/53\/9\/093020","article-title":"Particle transport analysis of the density build-up after the L\u2013H transition in ASDEX Upgrade","volume":"53","author":"Willensdorfer","year":"2013","journal-title":"Nucl. Fusion"},{"key":"mlstac3ffabib46","article-title":"Halomod: python package for dealing with the halo model","author":"Murray","year":"2017"},{"key":"mlstac3ffabib47","doi-asserted-by":"publisher","first-page":"46","DOI":"10.1088\/0004-637X\/728\/1\/46","article-title":"Galaxy clustering in the newfirm medium band survey: the relationship between stellar mass and dark matter halo mass at 1<z<2","volume":"728","author":"Wake","year":"2011","journal-title":"Astrophys. J."},{"key":"mlstac3ffabib48","doi-asserted-by":"publisher","first-page":"2929","DOI":"10.5194\/gmd-8-2929-2015","article-title":"MOPS-1.0: modelling the regulation of the global oceanic nitrogen budget by marine biogeochemical processes","volume":"8","author":"Kriest","year":"2015","journal-title":"Geosci. Model Dev."},{"key":"mlstac3ffabib49","doi-asserted-by":"publisher","DOI":"10.5281\/zenodo.1246300","article-title":"samarkhatiwala\/tmm: version 2.0 of the transport matrix method software","author":"Khatiwala","year":"2018"},{"key":"mlstac3ffabib50","doi-asserted-by":"publisher","first-page":"5753","DOI":"10.1029\/96JC02775","article-title":"A finite-volume, incompressible Navier Stokes model for studies of the ocean on parallel computers","volume":"102","author":"Marshall","year":"1997","journal-title":"J. Geophys. Res. Oceans"},{"key":"mlstac3ffabib51","doi-asserted-by":"publisher","first-page":"127","DOI":"10.5194\/gmd-10-127-2017","article-title":"Calibrating a global three-dimensional biogeochemical ocean model (MOPS-1.0)","volume":"10","author":"Kriest","year":"2017","journal-title":"Geosci. Model Dev."}],"container-title":["Machine Learning: Science and Technology"],"original-title":[],"link":[{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/ac3ffa","content-type":"text\/html","content-version":"am","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/ac3ffa\/pdf","content-type":"application\/pdf","content-version":"am","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/ac3ffa","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/ac3ffa\/pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/ac3ffa\/pdf","content-type":"application\/pdf","content-version":"am","intended-application":"syndication"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/ac3ffa\/pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"syndication"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/ac3ffa\/pdf","content-type":"application\/pdf","content-version":"am","intended-application":"similarity-checking"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/ac3ffa\/pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2021,12,27]],"date-time":"2021-12-27T16:47:34Z","timestamp":1640623654000},"score":1,"resource":{"primary":{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/ac3ffa"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2021,12,27]]},"references-count":51,"journal-issue":{"issue":"1","published-online":{"date-parts":[[2021,12,27]]},"published-print":{"date-parts":[[2022,3,1]]}},"URL":"https:\/\/doi.org\/10.1088\/2632-2153\/ac3ffa","relation":{},"ISSN":["2632-2153"],"issn-type":[{"value":"2632-2153","type":"electronic"}],"subject":[],"published":{"date-parts":[[2021,12,27]]},"assertion":[{"value":"Building high accuracy emulators for scientific simulations with deep neural architecture search","name":"article_title","label":"Article Title"},{"value":"Machine Learning: Science and Technology","name":"journal_title","label":"Journal Title"},{"value":"paper","name":"article_type","label":"Article Type"},{"value":"\u00a9 2021 The Author(s). Published by IOP Publishing Ltd","name":"copyright_information","label":"Copyright Information"},{"value":"2021-04-21","name":"date_received","label":"Date Received","group":{"name":"publication_dates","label":"Publication dates"}},{"value":"2021-12-03","name":"date_accepted","label":"Date Accepted","group":{"name":"publication_dates","label":"Publication dates"}},{"value":"2021-12-27","name":"date_epub","label":"Online publication date","group":{"name":"publication_dates","label":"Publication dates"}}]}}