{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,8,28]],"date-time":"2026-08-28T00:40:58Z","timestamp":1787877658347,"version":"build-2784847793"},"reference-count":42,"publisher":"American Mathematical Society (AMS)","issue":"359","license":[{"start":{"date-parts":[[2026,4,1]],"date-time":"2026-04-01T00:00:00Z","timestamp":1775001600000},"content-version":"am","delay-in-days":365,"URL":"https:\/\/www.ams.org\/publications\/copyright-and-permissions"}],"funder":[{"DOI":"10.13039\/501100001809","name":"National Natural Science Foundation of China","doi-asserted-by":"publisher","award":["12288101"],"award-info":[{"award-number":["12288101"]}],"id":[{"id":"10.13039\/501100001809","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["Math. Comp."],"abstract":"<p>We propose a hybrid iterative method based on multiple-input operator network for partial differential equations, which combines the traditional numerical iterative solver and the recent powerful machine learning method of neural operator, and further systematically analyzes its theoretical properties, including the convergence condition, the spectral behavior, as well as the convergence rate, in terms of the errors of the discretization and the model inference. We show the theoretical results for the frequently-used smoothers, i.e. Richardson (damped Jacobi) and Gauss-Seidel. We give an upper bound of the convergence rate of the hybrid method w.r.t. the model correction period, which indicates a minimum point to make the hybrid iteration converge fastest. Several numerical examples including the hybrid Richardson (Gauss-Seidel) iteration for the 1-d (2-d) Poisson equation are presented to verify our theoretical results, and also reflect an excellent acceleration effect. As a meshless acceleration method, it is provided with enormous potentials for practice applications.<\/p>","DOI":"10.1090\/mcom\/4086","type":"journal-article","created":{"date-parts":[[2025,3,26]],"date-time":"2025-03-26T14:23:22Z","timestamp":1742999002000},"page":"1327-1359","source":"Crossref","is-referenced-by-count":10,"title":["A hybrid iterative method based on MIONet for PDEs: Theory and numerical examples"],"prefix":"10.1090","volume":"95","author":[{"given":"Jun","family":"Hu","sequence":"first","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]},{"given":"Pengzhan","family":"Jin","sequence":"additional","affiliation":[],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"14","published-online":{"date-parts":[[2025,4,1]]},"reference":[{"issue":"138","key":"1","doi-asserted-by":"publisher","first-page":"333","DOI":"10.2307\/2006422","article-title":"Multi-level adaptive solutions to boundary-value problems","volume":"31","author":"Brandt, Achi","year":"1977","journal-title":"Math. Comp.","ISSN":"https:\/\/id.crossref.org\/issn\/0025-5718","issn-type":"print"},{"issue":"1-4","key":"2","doi-asserted-by":"publisher","first-page":"23","DOI":"10.1016\/0096-3003(86)90095-0","article-title":"Algebraic multigrid theory: the symmetric case","volume":"19","author":"Brandt, Achi","year":"1986","journal-title":"Appl. Math. Comput.","ISSN":"https:\/\/id.crossref.org\/issn\/0096-3003","issn-type":"print"},{"key":"3","doi-asserted-by":"crossref","unstructured":"S. C. Brenner, The Mathematical Theory of Finite Element Methods, Springer, 2008.","DOI":"10.1007\/978-0-387-75934-0"},{"key":"4","unstructured":"W. E, C. Ma, and L. Wu, Barron spaces and the compositional function spaces for neural network models, Preprint,  arXiv:1906.08039, 2019."},{"issue":"1","key":"5","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":"E, Weinan","year":"2018","journal-title":"Commun. Math. Stat.","ISSN":"https:\/\/id.crossref.org\/issn\/2194-6701","issn-type":"print"},{"key":"6","isbn-type":"print","doi-asserted-by":"publisher","first-page":"713","DOI":"10.1016\/S1570-8659(00)07005-8","article-title":"Finite volume methods","author":"Eymard, Robert","year":"2000","ISBN":"https:\/\/id.crossref.org\/isbn\/0444503501"},{"key":"7","unstructured":"G. Gupta, X. Xiao, and P. Bogdan, Multiwavelet-based operator learning for differential equations, Adv. Neural Inf. Process. Syst. 34 (2021), 24048\u201324062."},{"key":"8","unstructured":"W. Hackbusch, Multi-Grid Methods and Applications, vol. 4, Springer Science & Business Media, 2013."},{"key":"9","unstructured":"J. He, X. Liu, and J. Xu, MgNO: efficient parameterization of linear operators via multigrid, Preprint,  arXiv:2310.19809, 2023."},{"key":"10","unstructured":"J. Hu and P. Jin, Experimental observation on a low-rank tensor model for eigenvalue problems, Preprint,  arXiv:2302.00538, 2023."},{"key":"11","doi-asserted-by":"publisher","first-page":"109675","DOI":"10.1016\/j.jcp.2020.109675","article-title":"Int-Deep: a deep learning initialized iterative method for nonlinear problems","volume":"419","author":"Huang, Jianguo","year":"2020","journal-title":"J. Comput. Phys.","ISSN":"https:\/\/id.crossref.org\/issn\/0021-9991","issn-type":"print"},{"key":"12","doi-asserted-by":"crossref","unstructured":"Z. Jiang, M. Zhu, D. Li, Q. Li, Y. O. Yuan, and L. Lu, Fourier-MIONet: Fourier-enhanced multiple-input neural operators for multiphase modeling of geological carbon sequestration, Preprint,  arXiv:2303.04778, 2023.","DOI":"10.2139\/ssrn.4408632"},{"issue":"6","key":"13","doi-asserted-by":"publisher","first-page":"A3490--A3514","DOI":"10.1137\/22M1477751","article-title":"MIONet: learning multiple-input operators via tensor product","volume":"44","author":"Jin, Pengzhan","year":"2022","journal-title":"SIAM J. Sci. Comput.","ISSN":"https:\/\/id.crossref.org\/issn\/1064-8275","issn-type":"print"},{"key":"14","doi-asserted-by":"crossref","unstructured":"G. E. Karniadakis, I. G. Kevrekidis, L. Lu, P. Perdikaris, S. Wang, and L. Yang, Physics-informed machine learning, Nat. Rev. Phys. 3 (2021), no. 6, 422\u2013440.","DOI":"10.1038\/s42254-021-00314-5"},{"key":"15","first-page":"Paper No. [290], 76","article-title":"On universal approximation and error bounds for Fourier neural operators","volume":"22","author":"Kovachki, Nikola","year":"2021","journal-title":"J. Mach. Learn. Res.","ISSN":"https:\/\/id.crossref.org\/issn\/1532-4435","issn-type":"print"},{"issue":"1","key":"16","doi-asserted-by":"publisher","first-page":"tnac001, 141","DOI":"10.1093\/imatrm\/tnac001","article-title":"Error estimates for DeepONets: a deep learning framework in infinite dimensions","volume":"6","author":"Lanthaler, Samuel","year":"2022","journal-title":"Trans. Math. Appl."},{"key":"17","unstructured":"Z. Li, N. Kovachki, K. Azizzadenesheli, B. Liu, K. Bhattacharya, A. Stuart, and A. Anandkumar, Fourier neural operator for parametric partial differential equations, Preprint,  arXiv:2010.08895, 2020."},{"key":"18","unstructured":"Z. Li, N. Kovachki, K. Azizzadenesheli, B. Liu, K. Bhattacharya, A. Stuart, and A. Anandkumar, Neural operator: graph kernel network for partial differential equations, Preprint,  arXiv:2003.03485, 2020."},{"key":"19","unstructured":"L. Lu, P. Jin, and G. E. Karniadakis, DeepONet: learning nonlinear operators for identifying differential equations based on the universal approximation theorem of operators, Preprint,  arXiv:1910.03193, 2019."},{"key":"20","doi-asserted-by":"crossref","unstructured":"L. Lu, P. Jin, G. Pang, Z. Zhang, and G. E. Karniadakis, Learning nonlinear operators via DeepONet based on the universal approximation theorem of operators, Nat. Mach. Intell. 3 (2021), no. 3, 218\u2013229.","DOI":"10.1038\/s42256-021-00302-5"},{"issue":"6","key":"21","doi-asserted-by":"publisher","first-page":"B1105--B1132","DOI":"10.1137\/21M1397908","article-title":"Physics-informed neural networks with hard constraints for inverse design","volume":"43","author":"Lu, Lu","year":"2021","journal-title":"SIAM J. Sci. Comput.","ISSN":"https:\/\/id.crossref.org\/issn\/1064-8275","issn-type":"print"},{"issue":"1","key":"22","doi-asserted-by":"publisher","first-page":"39","DOI":"10.1016\/0096-3003(88)90063-X","article-title":"Algebraic study of multigrid methods for symmetric, definite problems","volume":"25","author":"Mandel, Jan","year":"1988","journal-title":"Appl. Math. Comput.","ISSN":"https:\/\/id.crossref.org\/issn\/0096-3003","issn-type":"print"},{"issue":"3","key":"23","doi-asserted-by":"publisher","first-page":"1513","DOI":"10.1137\/21M1465718","article-title":"Exponential convergence of deep operator networks for elliptic partial differential equations","volume":"61","author":"Marcati, Carlo","year":"2023","journal-title":"SIAM J. Numer. Anal.","ISSN":"https:\/\/id.crossref.org\/issn\/0036-1429","issn-type":"print"},{"key":"24","unstructured":"N. Rahaman, A. Baratin, D. Arpit, F. Draxler, M. Lin, F. Hamprecht, Y. Bengio, and A. Courville, On the spectral bias of neural networks, International Conference on Machine Learning, PMLR, 2019, pp. 5301\u20135310."},{"key":"25","unstructured":"M. A. Rahman, Z. E. Ross, and K. Azizzadenesheli, U-no: U-shaped neural operators, Preprint,  arXiv:2204.11127, 2022."},{"key":"26","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, M.","year":"2019","journal-title":"J. Comput. Phys.","ISSN":"https:\/\/id.crossref.org\/issn\/0021-9991","issn-type":"print"},{"key":"27","unstructured":"B. Raoni\u0107, R. Molinaro, T. Rohner, S. Mishra, and E. de Bezenac, Convolutional neural operators, Preprint,  arXiv:2302.01178, 2023."},{"key":"28","isbn-type":"print","first-page":"73","article-title":"Algebraic multigrid","author":"Ruge, J. W.","year":"1987","ISBN":"https:\/\/id.crossref.org\/isbn\/0898712149"},{"key":"29","series-title":"Springer Series in Computational Mathematics","isbn-type":"print","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-540-71041-7","volume-title":"Spectral methods","volume":"41","author":"Shen, Jie","year":"2011","ISBN":"https:\/\/id.crossref.org\/isbn\/9783540710400"},{"key":"30","doi-asserted-by":"publisher","first-page":"1","DOI":"10.1016\/j.acha.2021.12.005","article-title":"High-order approximation rates for shallow neural networks with cosine and \ud835\udc45\ud835\udc52\ud835\udc3f\ud835\udc48^{\ud835\udc58} activation functions","volume":"58","author":"Siegel, Jonathan W.","year":"2022","journal-title":"Appl. Comput. Harmon. Anal.","ISSN":"https:\/\/id.crossref.org\/issn\/1063-5203","issn-type":"print"},{"key":"31","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, Justin","year":"2018","journal-title":"J. Comput. Phys.","ISSN":"https:\/\/id.crossref.org\/issn\/0021-9991","issn-type":"print"},{"key":"32","series-title":"Oxford Applied Mathematics and Computing Science Series","isbn-type":"print","volume-title":"Numerical solution of partial differential equations","author":"Smith, G. D.","year":"1985","ISBN":"https:\/\/id.crossref.org\/isbn\/0198596413","edition":"3"},{"key":"33","isbn-type":"print","doi-asserted-by":"crossref","DOI":"10.1137\/1.9781611977165","volume-title":"Numerical linear algebra","author":"Trefethen, Lloyd N.","year":"[2022] \\copyright2022","ISBN":"https:\/\/id.crossref.org\/isbn\/9781611977158"},{"key":"34","isbn-type":"print","volume-title":"Multigrid","author":"Trottenberg, U.","year":"2001","ISBN":"https:\/\/id.crossref.org\/isbn\/012701070X"},{"key":"35","doi-asserted-by":"crossref","unstructured":"S. Wang, H. Wang, and P. Perdikaris, Learning the solution operator of parametric partial differential equations with physics-informed DeepONets, Sci. Adv. 7 (2021), no. 40, eabi8605.","DOI":"10.1126\/sciadv.abi8605"},{"issue":"6","key":"36","doi-asserted-by":"publisher","first-page":"1714","DOI":"10.4208\/jcm.2307-m2022-0233","article-title":"Tensor neural network and its numerical integration","volume":"42","author":"Wang, Yifan","year":"2024","journal-title":"J. Comput. Math.","ISSN":"https:\/\/id.crossref.org\/issn\/0254-9409","issn-type":"print"},{"key":"37","unstructured":"H. Wu, T. Hu, H. Luo, J. Wang, and M. Long, Solving high-dimensional PDEs with latent spectral models, Preprint,  arXiv:2301.12664, 2023."},{"key":"38","unstructured":"S. Xiao, P. Jin, and Y. Tang, A deformation-based framework for learning solution mappings of PDEs defined on varying domains, Preprint,  arXiv:2412.01379, 2024."},{"key":"39","volume-title":"Theory of multilevel methods","author":"Xu, Jinchao","year":"1989"},{"key":"40","doi-asserted-by":"publisher","first-page":"591","DOI":"10.1017\/S0962492917000083","article-title":"Algebraic multigrid methods","volume":"26","author":"Xu, Jinchao","year":"2017","journal-title":"Acta Numer.","ISSN":"https:\/\/id.crossref.org\/issn\/0962-4929","issn-type":"print"},{"issue":"5","key":"41","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, Zhi-Qin John","year":"2020","journal-title":"Commun. Comput. Phys.","ISSN":"https:\/\/id.crossref.org\/issn\/1815-2406","issn-type":"print"},{"key":"42","unstructured":"E. Zhang, A. Kahana, E. Turkel, R. Ranade, J. Pathak, and G. E. Karniadakis, A hybrid iterative numerical transferable solver (HINTS) for PDEs based on deep operator network and relaxation methods, Preprint,  arXiv:2208.13273, 2022."}],"container-title":["Mathematics of Computation"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.ams.org\/mcom\/2026-95-359\/S0025-5718-2025-04086-9\/mcom4086_AM.pdf","content-type":"application\/pdf","content-version":"am","intended-application":"syndication"},{"URL":"https:\/\/www.ams.org\/mcom\/2026-95-359\/S0025-5718-2025-04086-9\/S0025-5718-2025-04086-9.pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2026,4,22]],"date-time":"2026-04-22T06:01:41Z","timestamp":1776837701000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.ams.org\/mcom\/2026-95-359\/S0025-5718-2025-04086-9\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,4,1]]},"references-count":42,"journal-issue":{"issue":"359","published-print":{"date-parts":[[2026,5]]}},"alternative-id":["S0025-5718-2025-04086-9"],"URL":"https:\/\/doi.org\/10.1090\/mcom\/4086","archive":["CLOCKSS","Portico"],"relation":{},"ISSN":["1088-6842","0025-5718"],"issn-type":[{"value":"1088-6842","type":"electronic"},{"value":"0025-5718","type":"print"}],"subject":[],"published":{"date-parts":[[2025,4,1]]}}}