{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2025,2,22]],"date-time":"2025-02-22T00:00:27Z","timestamp":1740182427329,"version":"3.37.3"},"reference-count":44,"publisher":"IOP Publishing","issue":"1","license":[{"start":{"date-parts":[[2025,2,10]],"date-time":"2025-02-10T00:00:00Z","timestamp":1739145600000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"},{"start":{"date-parts":[[2025,2,10]],"date-time":"2025-02-10T00:00:00Z","timestamp":1739145600000},"content-version":"tdm","delay-in-days":0,"URL":"https:\/\/iopscience.iop.org\/info\/page\/text-and-data-mining"}],"funder":[{"DOI":"10.13039\/501100001809","name":"National Natural Science Foundation of China","doi-asserted-by":"crossref","award":["12071481"],"award-info":[{"award-number":["12071481"]}],"id":[{"id":"10.13039\/501100001809","id-type":"DOI","asserted-by":"crossref"}]}],"content-domain":{"domain":["iopscience.iop.org"],"crossmark-restriction":false},"short-container-title":["Mach. Learn.: Sci. Technol."],"published-print":{"date-parts":[[2025,3,31]]},"abstract":"<jats:title>Abstract<\/jats:title>\n               <jats:p>Ordinary deep neural network (DNN)-based methods frequently encounter difficulties when tackling multiscale and high-frequency partial differential equations. To overcome these obstacles and improve computational accuracy and efficiency, this paper presents the Randomized Radial Basis Function Neural Network (RRNN), an innovative approach explicitly crafted for solving multiscale elliptic equations. The RRNN method commences by decomposing the computational domain into non-overlapping subdomains. Within each subdomain, the solution to the localized subproblem is approximated by a RRNN with a Gaussian kernel. This network is distinguished by the random assignment of width and center coefficients for its activation functions, thereby rendering the training process focused solely on determining the weight coefficients of the output layer. For each subproblem, similar to the Petrov\u2013Galerkin finite element method, a linear system will be formulated on the foundation of a weak formulation. Subsequently, a selection of collocation points is stochastically sampled at the boundaries of the subdomain, ensuring the satisfaction of <jats:italic>C<\/jats:italic>\n                  <jats:sup>0<\/jats:sup> and <jats:italic>C<\/jats:italic>\n                  <jats:sup>1<\/jats:sup> continuity and boundary conditions to couple these localized solutions. The network is ultimately trained using the least squares method to ascertain the output layer weights. To validate the RRNN method\u2019s effectiveness, an extensive array of numerical experiments has been executed. The RRNN is firstly compared with a variety of DNN methods based on gradient descent optimization. The comparative analysis demonstrates the RRNN\u2019s superior performance with respect to computational accuracy and training time. Furthermore, it is contrasted with to local extreme learning machine method, which also utilizes domain decomposition and the least squares method. The comparative findings suggest that the RRNN method can attain enhanced accuracy at a comparable computational cost, particularly pronounced in scenarios with a smaller scale ratio <jats:italic>\u025b<\/jats:italic>.<\/jats:p>","DOI":"10.1088\/2632-2153\/ad979c","type":"journal-article","created":{"date-parts":[[2024,11,26]],"date-time":"2024-11-26T23:06:21Z","timestamp":1732662381000},"page":"015033","update-policy":"https:\/\/doi.org\/10.1088\/crossmark-policy","source":"Crossref","is-referenced-by-count":0,"title":["Randomized radial basis function neural network for solving multiscale elliptic equations"],"prefix":"10.1088","volume":"6","author":[{"ORCID":"https:\/\/orcid.org\/0009-0003-0719-3140","authenticated-orcid":true,"given":"Yuhang","family":"Wu","sequence":"first","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Ziyuan","family":"Liu","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Wenjun","family":"Sun","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Xu","family":"Qian","sequence":"additional","affiliation":[],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"266","published-online":{"date-parts":[[2025,2,10]]},"reference":[{"key":"mlstad979cbib1","doi-asserted-by":"publisher","first-page":"63","DOI":"10.1007\/s102550200004","article-title":"Multiscale domain decomposition methods for elliptic problems with high aspect ratios","volume":"18","author":"Aarnes","year":"2002","journal-title":"Acta Math. Appl. Sin."},{"key":"mlstad979cbib2","doi-asserted-by":"publisher","DOI":"10.1016\/j.cma.2023.116290","article-title":"Physics-informed radial basis network (PIRBN): a local approximating neural network for solving nonlinear partial differential equations","volume":"415","author":"Bai","year":"2023","journal-title":"Comput. Methods Appl. Mech. Eng."},{"key":"mlstad979cbib3","doi-asserted-by":"publisher","first-page":"45","DOI":"10.3389\/fmolb.2020.00045","article-title":"MiMiC: multiscale modeling in computational chemistry","volume":"7","author":"Bolnykh","year":"2020","journal-title":"Front. Mol. Biosci."},{"year":"2013","author":"Cai","key":"mlstad979cbib4"},{"key":"mlstad979cbib5","doi-asserted-by":"publisher","DOI":"10.1016\/j.jcp.2023.112481","article-title":"A multiscale computational framework using active learning to model complex suspension flows","volume":"493","author":"Chang","year":"2023","journal-title":"J. Comput. Phys."},{"key":"mlstad979cbib6","doi-asserted-by":"publisher","first-page":"268","DOI":"10.4208\/jml.220726","article-title":"Bridging traditional and machine learning-based algorithms for solving PDEs: the random feature method","volume":"1","author":"Chen","year":"2022","journal-title":"J. Mach. Learn."},{"key":"mlstad979cbib7","doi-asserted-by":"publisher","first-page":"435","DOI":"10.4208\/eajam.2023-065.050423","article-title":"The random feature method for time-dependent problems","volume":"13","author":"Chen","year":"2023","journal-title":"East Asian J. Appl. Math."},{"key":"mlstad979cbib8","doi-asserted-by":"publisher","first-page":"541","DOI":"10.1090\/S0025-5718-02-01441-2","article-title":"A mixed multiscale finite method for elliptic problems with oscillating coefficients","volume":"72","author":"Chen","year":"2002","journal-title":"Math. Comput."},{"key":"mlstad979cbib9","doi-asserted-by":"publisher","first-page":"86","DOI":"10.1039\/c0ib00075b","article-title":"Multi-scale modelling and simulation in systems biology","volume":"3","author":"Dada","year":"2011","journal-title":"Int. Biol."},{"key":"mlstad979cbib10","doi-asserted-by":"publisher","DOI":"10.1016\/j.cma.2021.114129","article-title":"Local extreme learning machines and domain decomposition for solving linear and nonlinear partial differential equations","volume":"387","author":"Dong","year":"2021","journal-title":"Comput. Methods Appl. Mech. Eng."},{"key":"mlstad979cbib11","doi-asserted-by":"publisher","first-page":"540","DOI":"10.1137\/S0036142996298880","article-title":"Wavelet-based numerical homogenization","volume":"35","author":"Dorobantu","year":"1998","journal-title":"SIAM J. Numer. Anal."},{"key":"mlstad979cbib12","doi-asserted-by":"publisher","first-page":"96","DOI":"10.1016\/j.neucom.2019.12.099","article-title":"Physics informed extreme learning machine (pielm)-a rapid method for the numerical solution of partial differential equations","volume":"391","author":"Dwivedi","year":"2020","journal-title":"Neurocomputing"},{"key":"mlstad979cbib13","doi-asserted-by":"publisher","first-page":"17","DOI":"10.1016\/0045-7825(95)00812-F","article-title":"Multigrid method for a periodic heterogeneous medium, Part I: Multiscale modeling and quality in multi-dimensional case","volume":"126","author":"Fish","year":"1995","journal-title":"Comput. Methods Appl. Mech. Eng."},{"key":"mlstad979cbib14","doi-asserted-by":"publisher","first-page":"306","DOI":"10.1109\/72.839002","article-title":"Selecting radial basis function network centers with recursive orthogonal least squares training","volume":"11","author":"Gomm","year":"2000","journal-title":"IEEE Trans. Neural Netw."},{"key":"mlstad979cbib15","doi-asserted-by":"publisher","first-page":"169","DOI":"10.1006\/jcph.1997.5682","article-title":"A multiscale finite element method for elliptic problems in composite materials and porous media","volume":"134","author":"Hou","year":"1997","journal-title":"J. Comput. Phys."},{"key":"mlstad979cbib16","doi-asserted-by":"publisher","first-page":"185","DOI":"10.4310\/CMS.2004.v2.n2.a3","article-title":"Removing the cell resonance error in the multiscale finite element method via a Petrov\u2013Galerkin formulation","volume":"2","author":"Hou","year":"2004","journal-title":"Commun. Math. Sci."},{"key":"mlstad979cbib17","doi-asserted-by":"publisher","first-page":"489","DOI":"10.1016\/j.neucom.2005.12.126","article-title":"Extreme learning machine: theory and applications","volume":"70","author":"Huang","year":"2006","journal-title":"Neurocomputing"},{"key":"mlstad979cbib18","doi-asserted-by":"publisher","first-page":"107","DOI":"10.1007\/s13042-011-0019-y","article-title":"Extreme learning machines: a survey","volume":"2","author":"Huang","year":"2011","journal-title":"Int. J. Mach. Learn. Cybern."},{"key":"mlstad979cbib19","doi-asserted-by":"publisher","first-page":"985","DOI":"10.1109\/IJCNN.2004.1380068","article-title":"Extreme learning machine: a new learning scheme of feedforward neural networks","volume":"vol 2","author":"Huang","year":"2004"},{"article-title":"Adam: a method for stochastic optimization","year":"2014","author":"Kingma","key":"mlstad979cbib20"},{"key":"mlstad979cbib21","doi-asserted-by":"publisher","first-page":"1886","DOI":"10.4208\/cicp.OA-2020-0187","article-title":"A multi-scale DNN algorithm for nonlinear elliptic equations with multiple scales","volume":"28","author":"Li","year":"2020","journal-title":"Commun. Comput. Phys."},{"key":"mlstad979cbib22","doi-asserted-by":"publisher","DOI":"10.1016\/j.jcp.2023.112242","article-title":"Subspace decomposition based DNN algorithm for elliptic type multiscale PDEs","volume":"488","author":"Li","year":"2023","journal-title":"J. Comput. Phys."},{"key":"mlstad979cbib23","doi-asserted-by":"publisher","first-page":"141","DOI":"10.11113\/matematika.v24.n.536","article-title":"An improved fast training algorithm for RBF networks using symmetry-based fuzzy C-means clustering","volume":"24","author":"Lim","year":"2008","journal-title":"Matematika"},{"key":"mlstad979cbib24","doi-asserted-by":"publisher","first-page":"66","DOI":"10.1038\/s43246-023-00391-2","article-title":"Multiscale mechanics and molecular dynamics simulations of the durability of fiber-reinforced polymer composites","volume":"4","author":"Lin","year":"2023","journal-title":"Commun. Mater."},{"key":"mlstad979cbib25","doi-asserted-by":"publisher","first-page":"1970","DOI":"10.4208\/cicp.OA-2020-0179","article-title":"Multi-scale deep neural network (mscalednn) for solving poisson-boltzmann equation in complex domains","volume":"28","author":"Liu","year":"2020","journal-title":"Commun. Comput. Phys."},{"key":"mlstad979cbib26","doi-asserted-by":"publisher","first-page":"133","DOI":"10.1016\/S0955-7997(01)00092-3","article-title":"Mesh-free radial basis function network methods with domain decomposition for approximation of functions and numerical solution of Poisson\u2019s equations","volume":"26","author":"Mai-Duy","year":"2002","journal-title":"Eng. Anal. Bound. Elem."},{"key":"mlstad979cbib27","doi-asserted-by":"publisher","first-page":"2583","DOI":"10.1090\/S0025-5718-2014-02868-8","article-title":"Localization of elliptic multiscale problems","volume":"83","author":"Malqvist","year":"2014","journal-title":"Math. Comput."},{"key":"mlstad979cbib28","doi-asserted-by":"publisher","first-page":"167","DOI":"10.1016\/j.neucom.2007.08.002","article-title":"Improvement of learning algorithms for RBF neural networks in a helicopter sound identification system","volume":"71","author":"Montazer","year":"2007","journal-title":"Neurocomputing"},{"key":"mlstad979cbib29","doi-asserted-by":"publisher","first-page":"281","DOI":"10.1162\/neco.1989.1.2.281","article-title":"Fast learning in networks of locally tuned process units","volume":"1","author":"Moody","year":"1989","journal-title":"Neural Comput."},{"key":"mlstad979cbib30","doi-asserted-by":"publisher","first-page":"62","DOI":"10.1007\/s10444-023-10065-9","article-title":"Finite basis physics-informed neural networks (FBPINNs): a scalable domain decomposition approach for solving differential equations","volume":"49","author":"Moseley","year":"2023","journal-title":"Adv. Comput. Math."},{"key":"mlstad979cbib31","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1137\/060670420","article-title":"Homogenization of parabolic equations with a continuum of space and time scales","volume":"46","author":"Owhadi","year":"2008","journal-title":"SIAM J. Numer. Anal."},{"key":"mlstad979cbib32","doi-asserted-by":"publisher","first-page":"453","DOI":"10.1146\/annurev.fluid.38.050304.092133","article-title":"Large-eddy simulation of turbulent combustion","volume":"38","author":"Pitsch","year":"2006","journal-title":"Annu. Rev. Fluid Mech."},{"key":"mlstad979cbib33","doi-asserted-by":"publisher","first-page":"169","DOI":"10.1007\/BF00195855","article-title":"Networks and the best approximation","volume":"63","author":"Poggio","year":"1990","journal-title":"Biol. Cybern."},{"key":"mlstad979cbib34","doi-asserted-by":"publisher","first-page":"697","DOI":"10.1016\/j.matcom.2022.10.018","article-title":"Solving partial differential equation based on extreme learning machine","volume":"205","author":"Quan","year":"2023","journal-title":"Math. Comput. SimulaT."},{"key":"mlstad979cbib35","doi-asserted-by":"publisher","first-page":"686","DOI":"10.1016\/j.jcp.2018.10.045","article-title":"Physics-informed neural networks: a deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations","volume":"378","author":"Raissi","year":"2019","journal-title":"J. Comput. Phys."},{"key":"mlstad979cbib36","doi-asserted-by":"publisher","DOI":"10.1016\/j.cnsns.2023.107518","article-title":"Randomized neural networks with Petrov\u2013Galerkin method for solving linear and nonlinear PDEs","volume":"127","author":"Shang","year":"2023","journal-title":"Commun. Nonlinear Sci."},{"key":"mlstad979cbib37","doi-asserted-by":"publisher","first-page":"1339","DOI":"10.1016\/j.jcp.2018.08.029","article-title":"DGM: a deep learning algorithm for solving partial differential equations","volume":"375","author":"Sirignano","year":"2018","journal-title":"J. Comput. Phys."},{"key":"mlstad979cbib38","doi-asserted-by":"publisher","first-page":"128","DOI":"10.1016\/j.camwa.2023.11.023","article-title":"Local randomized neural networks with discontinuous Galerkin methods for diffusive-viscous wave equation","volume":"154","author":"Sun","year":"2024","journal-title":"Commun. Nonlinear Sci. Numer. Simul."},{"key":"mlstad979cbib39","doi-asserted-by":"publisher","first-page":"75","DOI":"10.3868\/S140-DDD-023-0011-X","article-title":"Multiscale mathematical models for biological systems","volume":"18","author":"Sun","year":"2023","journal-title":"Front. Math. China"},{"key":"mlstad979cbib40","doi-asserted-by":"publisher","DOI":"10.1016\/j.jcp.2023.112452","article-title":"Solving multiscale elliptic problems by sparse radial basis function neural networks","volume":"492","author":"Wang","year":"2023","journal-title":"J. Comput. Phys."},{"article-title":"Adaptive local basis for elliptic problems with L \u221e-coefficients","year":"2017","author":"Weymuth","key":"mlstad979cbib41"},{"key":"mlstad979cbib42","doi-asserted-by":"publisher","first-page":"264","DOI":"10.1007\/978-3-030-36708-4_22","article-title":"Training behavior of deep neural network in frequency domain","author":"Xu","year":"2019"},{"key":"mlstad979cbib43","doi-asserted-by":"publisher","first-page":"1746","DOI":"10.4208\/cicp.OA-2020-0085","article-title":"Frequency principle: Fourier analysis sheds light on deep neural networks","volume":"28","author":"Xu","year":"2020","journal-title":"Commun. Comput. Phys."},{"key":"mlstad979cbib44","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1007\/s40304-018-0127-z","article-title":"The deep Ritz method: a deep learning-based numerical algorithm for solving variational problems","volume":"6","author":"Yang","year":"2018","journal-title":"Commun. Math. Stat."}],"container-title":["Machine Learning: Science and Technology"],"original-title":[],"link":[{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/ad979c","content-type":"text\/html","content-version":"am","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/ad979c\/pdf","content-type":"application\/pdf","content-version":"am","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/ad979c","content-type":"text\/html","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/ad979c\/pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"text-mining"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/ad979c\/pdf","content-type":"application\/pdf","content-version":"am","intended-application":"syndication"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/ad979c\/pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"syndication"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/ad979c\/pdf","content-type":"application\/pdf","content-version":"am","intended-application":"similarity-checking"},{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/ad979c\/pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,2,10]],"date-time":"2025-02-10T13:12:51Z","timestamp":1739193171000},"score":1,"resource":{"primary":{"URL":"https:\/\/iopscience.iop.org\/article\/10.1088\/2632-2153\/ad979c"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,2,10]]},"references-count":44,"journal-issue":{"issue":"1","published-online":{"date-parts":[[2025,2,10]]},"published-print":{"date-parts":[[2025,3,31]]}},"URL":"https:\/\/doi.org\/10.1088\/2632-2153\/ad979c","relation":{},"ISSN":["2632-2153"],"issn-type":[{"type":"electronic","value":"2632-2153"}],"subject":[],"published":{"date-parts":[[2025,2,10]]},"assertion":[{"value":"Randomized radial basis function neural network for solving multiscale elliptic equations","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 2025 The Author(s). Published by IOP Publishing Ltd","name":"copyright_information","label":"Copyright Information"},{"value":"2024-10-06","name":"date_received","label":"Date Received","group":{"name":"publication_dates","label":"Publication dates"}},{"value":"2024-11-26","name":"date_accepted","label":"Date Accepted","group":{"name":"publication_dates","label":"Publication dates"}},{"value":"2025-02-10","name":"date_epub","label":"Online publication date","group":{"name":"publication_dates","label":"Publication dates"}}]}}