{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2025,10,13]],"date-time":"2025-10-13T15:41:43Z","timestamp":1760370103786,"version":"build-2065373602"},"reference-count":32,"publisher":"MDPI AG","issue":"1","license":[{"start":{"date-parts":[[2025,1,20]],"date-time":"2025-01-20T00:00:00Z","timestamp":1737331200000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/creativecommons.org\/licenses\/by\/4.0\/"}],"funder":[{"DOI":"10.13039\/501100001809","name":"National Natural Science Foundation of China","doi-asserted-by":"publisher","award":["12161067"],"award-info":[{"award-number":["12161067"]}],"id":[{"id":"10.13039\/501100001809","id-type":"DOI","asserted-by":"publisher"}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["Axioms"],"abstract":"<jats:p>Deep learning methods using neural networks for solving partial differential equations (PDEs) have emerged as a new paradigm. However, many of these methods approximate solutions by optimizing loss functions, often encountering convergence issues and accuracy limitations. In this paper, we propose a novel deep learning approach that leverages the expressive power of neural networks to generate basis functions. These basis functions are then used to create trial solutions, which are optimized using the least-squares method to solve for coefficients in a system of linear equations. This method integrates the strengths of streaming PINNs and the traditional least-squares method, offering both flexibility and a high accuracy. We conducted numerical experiments to compare our method with the results of high-order finite difference schemes and several commonly used neural network methods (PINNs, lbPINNs, ELMs, and PIELMs). Thanks to the mesh-less feature of the neural network, it is particularly effective for complex geometries. The numerical results demonstrate that our method significantly enhances the accuracy of deep learning in solving PDEs, achieving error levels comparable to high-accuracy finite difference methods.<\/jats:p>","DOI":"10.3390\/axioms14010075","type":"journal-article","created":{"date-parts":[[2025,1,20]],"date-time":"2025-01-20T12:32:52Z","timestamp":1737376372000},"page":"75","update-policy":"https:\/\/doi.org\/10.3390\/mdpi_crossmark_policy","source":"Crossref","is-referenced-by-count":3,"title":["A Novel Neural Network-Based Approach Comparable to High-Precision Finite Difference Methods"],"prefix":"10.3390","volume":"14","author":[{"given":"Fanghua","family":"Pei","sequence":"first","affiliation":[{"name":"School of Mathematical Statistics, Ningxia University, Yinchuan 750021, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0003-4188-495X","authenticated-orcid":false,"given":"Fujun","family":"Cao","sequence":"additional","affiliation":[{"name":"School of Science, Inner Mongolia University of Science and Technology, Baotou 014010, China"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Yongbin","family":"Ge","sequence":"additional","affiliation":[{"name":"School of Mathematical Statistics, Ningxia University, Yinchuan 750021, China"},{"name":"School of Science, Dalian Minzu University, Dalian 116600, China"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"1968","published-online":{"date-parts":[[2025,1,20]]},"reference":[{"key":"ref_1","unstructured":"Ames, W.F. (2014). Numerical Methods for Partial Differential Equations, Academic Press."},{"key":"ref_2","unstructured":"Morton, K.W., and Mayers, D.F. (1995). Numerical Solution of Partial Differential Equations, Cambridge University Press."},{"key":"ref_3","unstructured":"Spotz, W.F., and Carey, G.F. (1995, January 5\u20139). High-order compact finite difference methods. Proceedings of the Preliminary Proceedings International Conference on Spectral and High Order Methods, Houston, TX, USA."},{"key":"ref_4","doi-asserted-by":"crossref","first-page":"107797","DOI":"10.1016\/j.cnsns.2023.107797","article-title":"Conservative higher-order finite difference scheme for the coupled nonlinear Schr\u00f6dinger equations","volume":"131","author":"Liu","year":"2024","journal-title":"Commun. Nonlinear Sci. Numer. Simul."},{"key":"ref_5","doi-asserted-by":"crossref","first-page":"100","DOI":"10.1016\/j.camwa.2024.05.028","article-title":"Numerical solutions of the nonlinear wave equations with energy-preserving sixth-order finite difference schemes","volume":"168","author":"Wang","year":"2024","journal-title":"Comput. Math. Appl."},{"key":"ref_6","doi-asserted-by":"crossref","first-page":"113361","DOI":"10.1016\/j.jcp.2024.113361","article-title":"A sixth-order finite difference method for the two-dimensional nonlinear advection diffusion reaction equation","volume":"517","author":"Liu","year":"2024","journal-title":"J. Comput. Phys."},{"key":"ref_7","doi-asserted-by":"crossref","first-page":"952","DOI":"10.1016\/j.jcp.2006.06.001","article-title":"High-order compact exponential finite difference methods for convection\u2013diffusion type problems","volume":"220","author":"Tian","year":"2007","journal-title":"J. Comput. Phys."},{"key":"ref_8","doi-asserted-by":"crossref","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":"ref_9","doi-asserted-by":"crossref","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":"Yu","year":"2018","journal-title":"Commun. Math. Stat."},{"key":"ref_10","doi-asserted-by":"crossref","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":"ref_11","unstructured":"Huang, G.B., Zhu, Q.Y., and Siew, C.K. (2004, January 25\u201329). Extreme learning machine: A new learning scheme of feedforward neural networks. Proceedings of the 2004 IEEE International Joint Conference on Neural Networks, Budapest, Hungary."},{"key":"ref_12","doi-asserted-by":"crossref","first-page":"11","DOI":"10.1016\/j.neucom.2022.05.015","article-title":"Self-adaptive loss balanced Physics-informed neural networks","volume":"496","author":"Xiang","year":"2022","journal-title":"Neurocomputing"},{"key":"ref_13","doi-asserted-by":"crossref","first-page":"88","DOI":"10.1007\/s10915-022-01939-z","article-title":"Scientific machine learning through physics\u2013informed neural networks: Where we are and what\u2019s next","volume":"92","author":"Cuomo","year":"2022","journal-title":"J. Sci. Comput."},{"key":"ref_14","doi-asserted-by":"crossref","unstructured":"Markidis, S. (2021). The old and the new: Can physics-informed deep-learning replace traditional linear solvers?. Front. Big Data, 4.","DOI":"10.3389\/fdata.2021.669097"},{"key":"ref_15","doi-asserted-by":"crossref","first-page":"106998","DOI":"10.1016\/j.neunet.2024.106998","article-title":"An extrapolation-driven network architecture for physics-informed deep learning","volume":"183","author":"Wang","year":"2025","journal-title":"Neural Netw."},{"key":"ref_16","unstructured":"Zeng, B., Wang, Q., Yan, M., Liu, Y., Chengze, R., Zhang, Y., Liu, H., Wang, Z., and Sun, H. (2024). PhyMPGN: Physics-encoded Message Passing Graph Network for spatiotemporal PDE systems. arXiv."},{"key":"ref_17","doi-asserted-by":"crossref","first-page":"110768","DOI":"10.1016\/j.jcp.2021.110768","article-title":"When and why PINNs fail to train: A neural tangent kernel perspective","volume":"449","author":"Wang","year":"2022","journal-title":"J. Comput. Phys."},{"key":"ref_18","doi-asserted-by":"crossref","unstructured":"Urb\u00e1n, J.F., Stefanou, P., and Pons, J.A. (2024). Unveiling the optimization process of Physics Informed Neural Networks: How accurate and competitive can PINNs be?. arXiv.","DOI":"10.1016\/j.jcp.2024.113656"},{"key":"ref_19","unstructured":"Rathore, P., Lei, W., Frangella, Z., Lu, L., and Udell, M. (2024). Challenges in training PINNs: A loss landscape perspective. arXiv."},{"key":"ref_20","doi-asserted-by":"crossref","first-page":"23943","DOI":"10.1109\/ACCESS.2024.3354058","article-title":"An improved method for Physics-Informed Neural Networks that accelerates convergence","volume":"12","author":"Yan","year":"2024","journal-title":"IEEE Access"},{"key":"ref_21","doi-asserted-by":"crossref","first-page":"112918","DOI":"10.1016\/j.jcp.2024.112918","article-title":"Correcting model misspecification in physics-informed neural networks (PINNs)","volume":"505","author":"Zou","year":"2024","journal-title":"J. Comput. Phys."},{"key":"ref_22","unstructured":"M\u00fcller, J., and Zeinhofer, M. (2004, January 25\u201329). Achieving high accuracy with PINNs via energy natural gradient descent. Proceedings of the International Conference on Machine Learning, Budapest, Hungary."},{"key":"ref_23","doi-asserted-by":"crossref","first-page":"11064","DOI":"10.1109\/TNNLS.2023.3247163","article-title":"Physics-informed neural networks with weighted losses by uncertainty evaluation for accurate and stable prediction of manufacturing systems","volume":"35","author":"Hua","year":"2024","journal-title":"IEEE Trans. Neural Netw. Learn. Syst."},{"key":"ref_24","doi-asserted-by":"crossref","first-page":"1","DOI":"10.1093\/imanum\/drab093","article-title":"Estimates on the generalization error of physics-informed neural networks for approximating PDEs","volume":"43","author":"Mishra","year":"2023","journal-title":"IMA J. Numer. Anal."},{"key":"ref_25","doi-asserted-by":"crossref","first-page":"96","DOI":"10.1016\/j.neucom.2019.12.099","article-title":"Physics informed extreme learning machine (pielm)\u2013a rapid method for the numerical solution of partial differential equations","volume":"391","author":"Dwivedi","year":"2020","journal-title":"Neurocomputing"},{"key":"ref_26","doi-asserted-by":"crossref","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. Simul."},{"key":"ref_27","doi-asserted-by":"crossref","first-page":"116578","DOI":"10.1016\/j.cma.2023.116578","article-title":"An extreme learning machine-based method for computational PDEs in higher dimensions","volume":"418","author":"Wang","year":"2024","journal-title":"Comput. Methods Appl. Mech. Eng."},{"key":"ref_28","doi-asserted-by":"crossref","first-page":"114129","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":"ref_29","unstructured":"Xu, Z., and Sheng, Z. (2024). Subspace method based on neural networks for solving the partial differential equation. arXiv."},{"key":"ref_30","first-page":"747","article-title":"High-precision finite-difference method for the nonlinear reaction-diffusion equation","volume":"36","author":"Sheng","year":"2023","journal-title":"Appl. Math."},{"key":"ref_31","doi-asserted-by":"crossref","first-page":"103120","DOI":"10.1016\/j.wavemoti.2023.103120","article-title":"Explicit high-order compact difference method for solving nonlinear hyperbolic equations with three types of boundary conditions","volume":"118","author":"Meng","year":"2023","journal-title":"Wave Motion"},{"key":"ref_32","first-page":"92","article-title":"A high precision compact difference method for solving the Helmholtz equation of variable coefficient","volume":"36","author":"Zhi","year":"2019","journal-title":"J. Chongqing Norm. Univ."}],"container-title":["Axioms"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/www.mdpi.com\/2075-1680\/14\/1\/75\/pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2025,10,8]],"date-time":"2025-10-08T10:32:24Z","timestamp":1759919544000},"score":1,"resource":{"primary":{"URL":"https:\/\/www.mdpi.com\/2075-1680\/14\/1\/75"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2025,1,20]]},"references-count":32,"journal-issue":{"issue":"1","published-online":{"date-parts":[[2025,1]]}},"alternative-id":["axioms14010075"],"URL":"https:\/\/doi.org\/10.3390\/axioms14010075","relation":{},"ISSN":["2075-1680"],"issn-type":[{"type":"electronic","value":"2075-1680"}],"subject":[],"published":{"date-parts":[[2025,1,20]]}}}