[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"doc-detail-117324-en":3,"doc-seo-117324-105":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},117324,1099514068035,"Ezra","https://ap-avatar.wpscdn.com/davatar_276721f389ce27ea32af1340a28f341c",8,"Research & Report","Machine learning for accuracy in density functional approximations","Machine learning techniques accelerate atomistic simulations and materials design in computational chemistry, while also enhancing the predictive power of efficient electronic-structure methods such as density functional theory. The work reviews recent progress applying machine learning to improve density functional and related approximations, aiming for chemical accuracy and correction of fundamental density functional errors. It discusses promises and challenges for building models transferable across chemistries and material classes, supported by examples tested beyond their training sets.","arXiv :2311 .00196v1 [physics .chem-ph] 1 Nov 2023  \nMachine learning for accuracy in density functional  \napproximations  \nJohannes Voss∗†  \nNovember 2, 2023  \nAbstract  \nMachine learning techniques have found their way into computational chemistry as indispensable tools to accelerate atomistic simulations and materials design. In addition, machine learning approaches hold the potential to boost the predictive power of computationally efficient electronic structure methods, such as density functional theory, to chemical accuracy and to correct for fundamental errors in density functional approaches. Here, recent progress in applying machine learning to improve the accuracy of density functional and related approximations is reviewed. Promises and challenges in devising machine learning models transferable between different chemistries and materials classes are discussed with the help of examples applying promising models to systems far outside their training sets.  \nKeywords: Machine learning, density functional theory, materials prediction,  \nexchange-correlation functional, self-interaction, electron delocalization   \n∗ SUNCAT Center for Interface Science and Catalysis, SLAC National Accelerator Laboratory, 2575 Sand Hill Road, Menlo Park, CA 94025, USA  \n†[vossj@slac.stanford.edu](vossj@slac.stanford.edu)  \nMachine learning techniques allow us, where benchmark data are available, to train electronic structure models that substantially increase the predictive power of density functional theory simulations of chemical reactions and structural and thermodynamic properties of gas, liquid, and solid phases. Not only can quantitative improvements be achieved, but also fundamental limitations of density functional approximations can be corrected for. Here, techniques, benchmark data, and challenges for devising transferable electronic structure machine learning models are reviewed.  \nFigure 1: Overview of ML approaches to increasing the accuracy of electronic structure predictions based on electronic or atomic structural features. Machine-learned XC functionals are trained on high-accuracy benchmark data to improve upon the predictive power of existing DFAs. Post-DFT and ∆-ML methods provide improved energetics on fixed DFT charge densities, and other ML approaches supplement the Kohn-Sham Hamiltonian with Hubbard and dispersion terms.  \n1 INTRODUCTION  \nMachine learning (ML) techniques play an increasingly important role in atomistic-scale simulations in computational chemistry and physics. 1–3 Major areas of research are the acceleration of materials discovery and extending computationally accessible time and length scales through accelerated simulations. Inter-atomic potentials represented by neural networks 4–7 or other ML regression techniques 8,9 enable accurate molecular dynamics simulations for system sizes and time scales well beyond what can be achieved with first-principles Hamiltonians. 10 When computation of the Born-Oppenheimer potential energy surface isnot required, ML approaches trained to map chemical composition and other not necessarily atomic structure sensitive features to system properties of interest are powerful methods for direct, approximate materials property predictions. 11–15 Such methods can furthermore be employed for inverse materials design, where molecules or materials compositions that could lead to a desired target metric are predicted. 16,17 These models and inter-atomic potentials are trained on high-throughput datasets generated with computationally affordable methods. Density functional theory (DFT) 18 is often the method of choice due to a favorable trade-off between computational complexity and accuracy for the prediction of the electronic structures of molecules and solids. 19,20 Some (minor or appreciable) loss in accuracy with respect to the DFT training data is typically tolerated with the advantage of significant speed up of the resulting ML methods over DFT simulations. At best, th","cbCais649SYhtsJU","https://ap.wps.com/l/cbCais649SYhtsJU","pdf",2960383,1,28,"English","en",105,"# Abstract\n# Introduction\n## Machine learning for materials discovery and property prediction\n## ML for improving DFT accuracy\n# Shortcomings of density functional approximations","[{\"question\":\"How does machine learning improve density functional theory predictions?\",\"answer\":\"Machine learning is used to train electronic-structure models that increase the predictive power of density functional theory and can correct fundamental limitations of density functional approximations. Approaches include machine-learned functionals, Hamiltonian corrections, and Δ-ML post-DFT corrections.\"},{\"question\":\"What categories of ML methods are discussed for improving DFT accuracy?\",\"answer\":\"The document groups methods into machine-learned exchange-correlation (XC) functionals, atomic-structure-dependent machine-learned Hamiltonian corrections, and Δ-ML approaches that learn corrections to apply to DFT results, with some methods overlapping multiple categories.\"},{\"question\":\"What challenges arise when applying transferable ML models to new chemical systems?\",\"answer\":\"Challenges include limited availability of accurate training data for challenging systems and transferability issues when models are applied outside their training distribution. Examples illustrate these problems for promising models tested beyond their training sets.\"}]","Machine learning for accuracy in density functional approximations | PDF",1785675184,71,{"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},"machine-learning-for-accuracy-in-density-functional-approximations","",{"@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/machine-learning-for-accuracy-in-density-functional-approximations/117324/",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","2026-08-02",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},"How does machine learning improve density functional theory predictions?","Question",{"text":75,"@type":76},"Machine learning is used to train electronic-structure models that increase the predictive power of density functional theory and can correct fundamental limitations of density functional approximations. Approaches include machine-learned functionals, Hamiltonian corrections, and Δ-ML post-DFT corrections.","Answer",{"name":78,"@type":73,"acceptedAnswer":79},"What categories of ML methods are discussed for improving DFT accuracy?",{"text":80,"@type":76},"The document groups methods into machine-learned exchange-correlation (XC) functionals, atomic-structure-dependent machine-learned Hamiltonian corrections, and Δ-ML approaches that learn corrections to apply to DFT results, with some methods overlapping multiple categories.",{"name":82,"@type":73,"acceptedAnswer":83},"What challenges arise when applying transferable ML models to new chemical systems?",{"text":84,"@type":76},"Challenges include limited availability of accurate training data for challenging systems and transferability issues when models are applied outside their training distribution. Examples illustrate these problems for promising models tested beyond their training sets.","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"]