[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-119497-en":3,"doc-seo-119497-105":30,"detail-sidebar-cat-0-en-105":91},{"code":4,"msg":5,"data":6},0,"success",{"doc_id":7,"user_id":8,"nickname":9,"user_avatar":10,"doc_module":4,"category_id":11,"category_name":12,"doc_title":13,"doc_description":14,"doc_content":15,"file_id":16,"file_url":17,"file_type":18,"file_size":19,"view_count":4,"is_deleted":4,"is_public":20,"is_downloadable":20,"audit_status":20,"page_count":21,"language":22,"language_code":23,"site_id":24,"html_lang":23,"table_of_contents":25,"faqs":26,"seo_title":27,"seo_description":14,"update_tm":28,"read_time":29},119497,962075114101,"Seraphina","https://ap-avatar.wpscdn.com/avatar/e000253a75eb197efd?x-image-process=image/resize,m_fixed,w_180,h_180&k=1780044092746381165",8,"Research & Report","Uncertainty in the era of machine learning for atomistic modeling - Perspective","Machine learning surrogate models enable atomistic modeling to explore larger, more complex systems with high accuracy and efficiency, yet their data-driven nature produces uncertainties that must be rigorously quantified and managed. This perspective surveys state-of-the-art uncertainty estimation approaches, from Bayesian frameworks to ensembling, and explains how they apply to atomistic modeling. It also analyzes how predictive accuracy, uncertainty behavior, training-data composition, acquisition strategies, transferability, and robustness interact, synthesizing current debates and limitations.","Digital  \nDiscovery  \nPERSPECTIVE  \nCite this: DOI: 10 .1039/d5dd00102a  \nReceived 11th March 2025  \nAccepted 5th June 2025  \nDOI: 10.1039/d5dd00102a[rsc.li/digitaldiscovery](rsc.li/digitaldiscovery)  \nUncertainty in the era of machine learning for atomistic modeling  \nFederico Grasselli,*ab Sanggyu Chong,  c Venkat Kapil,def Silvia Bonfantighand Kevin Rossi  *ij  \nThe widespread adoption of machine learning surrogate models has significantly improved the scale and complexity of systems and processes that can be explored accurately and efficiently using atomistic modeling. However, the inherently data-driven nature of machine learning models introduces uncertainties that must be quantified, understood, and effectively managed to ensure reliable predictions and conclusions. Building upon these premises, in this perspective, we first overview state-of-the-art uncertainty estimation methods, from Bayesian frameworks to ensembling techniques, and discuss their application in atomistic modeling. We then examine the interplay between model accuracy, uncertainty, training dataset composition, data acquisition strategies, model transferability, and robustness. In doing so, we synthesize insights from the existing literature and highlight areas of ongoing debate.  \n1 Introduction  \nTycho Brahe, 16th century Danish astronomer, is credited for the “great care he took in correcting his observations for instrumental errors”,1 introducing the concept of measurement-theory inconsistency in astronomy, thus turning it into an empirical science. Since then, the ability to assess instrument and model errors as well as quantify the uncertainty and con􀀁dence intervals when making predictions has become a pillar of the scienti􀀁c method and, in fact, discriminates between what is scienti􀀁c and what is not.  \nIn many cases, chemists and materials scientists draw conclusions based on incomplete or uncertain information, as it is o􀀁en the case when dealing with expensive, time-  \naDipartimento di Scienze Fisiche, Informatiche e Matematiche, Universit degli Studidi Modena e Reggio Emilia, 41125 Modena, Italy. E-mail: federico.grasselli@ [unimore.it](unimore.it)  \nbCNR NANO S3, 41125 Modena, Italy  \ncLaboratory of Computational Science and Modeling, Institute of Materials, ´Ecole Polytechnique Fdrale de Lausanne, 1015 Lausanne, Switzerland  \ndYusuf Hamied Department of Chemistry, University of Cambridge, Cambridge CB2 1EW, UK  \neDepartment of Physics and Astronomy, University College London, London, UK fThomas Young Centre, London Centre for Nanotechnology, University College London, London WC1E 6BT, UK  \ngCenter for Complexity and Biosystems, Department of Physics “Aldo Pontremoli”, University of Milan, Via Celoria 16, 20133 Milano, Italy  \nhNOMATEN Centre of Excellence, National Center for Nuclear Research, ul. A. Sołtana 7, 05-400 Swierk/Otwock, Poland  \niDepartment of Materials Science and Engineering, Del􀀁 University of Technology, Del􀀁, 2628 CD, The Netherlands. E-mail: k.r.rossi@tudel􀀁.nl  \njClimate Safety and Security Centre, Del􀀁 University of Technology, TU Del􀀁 the Hague Campus, The Hague, 2594 AC, The Netherlands  \nconsuming, o􀀁entimes noisy measurements. Uncertainty quanti􀀁cation (UQ) provides a framework for systematically incorporating uncertainty in this scienti􀀁c process, thereby enhancing the reliability, robustness, and applicability of experimental and theoretical results. In the context of materials science, chemistry, and condensed matter physics, researchers optimize materials and properties while accounting for uncertainties, variations, and errors in their measurements and theories (e.g., via replication and sensitivity analysis) . This improves the reliability and validity in the models of physical phenomena and design of novel materials and processes.  \nNowadays, machine learning and arti􀀁cial intelligence methods are emerging as a key tools for accelerating the design, engineering, characterization, and understanding of materials, m","cbCaidTpRly90s4Y","https://ap.wps.com/l/cbCaidTpRly90s4Y","pdf",2649581,1,22,"English","en",105,"# Introduction\n## Uncertainty quantification in scientific inference\n## Machine learning for atomistic modeling\n## Integrating UQ with surrogate models","[{\"question\":\"Why is uncertainty quantification important for machine learning atomistic models?\",\"answer\":\"Machine learning models are inherently data-driven, so they carry uncertainty that can affect the reliability of predictions and conclusions. Quantifying uncertainty supports robust decision-making and proper interpretation of model limitations.\"},{\"question\":\"Which uncertainty estimation approaches does the perspective review?\",\"answer\":\"It reviews state-of-the-art methods including Bayesian frameworks and ensembling techniques. These are discussed in terms of their application to atomistic modeling.\"},{\"question\":\"What factors influence the predictive reliability of machine learning models in atomistic modeling?\",\"answer\":\"The perspective examines the interplay between model accuracy and uncertainty, training dataset composition, data acquisition strategies, model transferability, and robustness. It links these factors to how models are trained and deployed.\"}]","Uncertainty in the era of machine learning for atomistic modeling - Perspective | PDF",1785724626,55,{"code":4,"msg":31,"data":32},"ok",{"site_id":24,"language":23,"slug":33,"title":13,"keywords":34,"description":14,"schema_data":35,"social_meta":86,"head_meta":88,"extra_data":90,"updated_unix":28},"uncertainty-in-the-era-of-machine-learning-for-atomistic-modeling-perspective","",{"@graph":36,"@context":85},[37,54,68],{"@type":38,"itemListElement":39},"BreadcrumbList",[40,44,48,51],{"item":41,"name":42,"@type":43,"position":20},"https://docshare.wps.com","Home","ListItem",{"item":45,"name":46,"@type":43,"position":47},"https://docshare.wps.com/document/","Document",2,{"item":49,"name":12,"@type":43,"position":50},"https://docshare.wps.com/document/research-report/",3,{"item":52,"name":13,"@type":43,"position":53},"https://docshare.wps.com/document/uncertainty-in-the-era-of-machine-learning-for-atomistic-modeling-perspective/119497/",4,{"url":52,"name":13,"@type":55,"author":56,"headline":13,"publisher":58,"fileFormat":61,"inLanguage":23,"description":14,"dateModified":62,"datePublished":62,"encodingFormat":61,"isAccessibleForFree":63,"interactionStatistic":64},"DigitalDocument",{"name":9,"@type":57},"Person",{"url":41,"name":59,"@type":60},"DocShare","Organization","application/pdf","2026-08-03",true,{"@type":65,"interactionType":66,"userInteractionCount":4},"InteractionCounter",{"@type":67},"ViewAction",{"@type":69,"mainEntity":70},"FAQPage",[71,77,81],{"name":72,"@type":73,"acceptedAnswer":74},"Why is uncertainty quantification important for machine learning atomistic models?","Question",{"text":75,"@type":76},"Machine learning models are inherently data-driven, so they carry uncertainty that can affect the reliability of predictions and conclusions. Quantifying uncertainty supports robust decision-making and proper interpretation of model limitations.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"Which uncertainty estimation approaches does the perspective review?",{"text":80,"@type":76},"It reviews state-of-the-art methods including Bayesian frameworks and ensembling techniques. These are discussed in terms of their application to atomistic modeling.",{"name":82,"@type":73,"acceptedAnswer":83},"What factors influence the predictive reliability of machine learning models in atomistic modeling?",{"text":84,"@type":76},"The perspective examines the interplay between model accuracy and uncertainty, training dataset composition, data acquisition strategies, model transferability, and robustness. It links these factors to how models are trained and deployed.","https://schema.org",{"og:url":52,"og:type":87,"og:title":13,"og:site_name":59,"og:description":14},"article",{"robots":89,"canonical":52},"index,follow",{"doc_id":7,"site_id":24},{"code":4,"msg":5,"data":92},[93,97,101,105,110,115,120,123,128,131,135],{"id":20,"doc_module":4,"doc_module_name":46,"category_name":94,"show_sort_weight":95,"slug":96},"Story & Novel",90,"story-novel",{"id":47,"doc_module":4,"doc_module_name":46,"category_name":98,"show_sort_weight":99,"slug":100},"Literature",80,"literature",{"id":53,"doc_module":4,"doc_module_name":46,"category_name":102,"show_sort_weight":103,"slug":104},"Exam",70,"exam",{"id":106,"doc_module":4,"doc_module_name":46,"category_name":107,"show_sort_weight":108,"slug":109},5,"Comic",60,"comic",{"id":111,"doc_module":4,"doc_module_name":46,"category_name":112,"show_sort_weight":113,"slug":114},6,"Technology",50,"technology",{"id":116,"doc_module":4,"doc_module_name":46,"category_name":117,"show_sort_weight":118,"slug":119},7,"Healthcare",40,"healthcare",{"id":11,"doc_module":4,"doc_module_name":46,"category_name":12,"show_sort_weight":121,"slug":122},30,"research-report",{"id":124,"doc_module":4,"doc_module_name":46,"category_name":125,"show_sort_weight":126,"slug":127},9,"Religion & Spirituality",20,"religion-spirituality",{"id":126,"doc_module":4,"doc_module_name":46,"category_name":129,"show_sort_weight":126,"slug":130},"World Cup","world-cup",{"id":132,"doc_module":4,"doc_module_name":46,"category_name":133,"show_sort_weight":132,"slug":134},10,"Lifestyle","lifestyle",{"id":136,"doc_module":4,"doc_module_name":46,"category_name":137,"show_sort_weight":106,"slug":138},19,"General","general"]