Weak baselines and reporting biases lead to overoptimism in machine learning for fluid-related partial differential equations

Weak baselines and reporting biases lead to overoptimism in machine learning for fluid-related partial differential equations

  • Randall, D. & Welser, C. The Irreproducibility Crisis of Modern Science: Causes, Consequences, and the Road to Reform (National Association of Scholars, 2018).

  • Ritchie, S. Science fictions: How Fraud, Bias, Negligence, and Hype Undermine the Search for Truth (Vintage, 2020).

  • Munafò, M. R. et al. A manifesto for reproducible science. Nat. Human Behav. 1, 1–9 (2017).

    Article 

    Google Scholar 

  • Ioannidis, J. P. A. Why most published research findings are false. PLoS Med. 2, e124 (2005).

    Article 

    Google Scholar 

  • Open Science Collaboration. Estimating the reproducibility of psychological science. Science 349, aac4716 (2015).

    Article 

    Google Scholar 

  • Prinz, F., Schlange, T. & Asadullah, K. Believe it or not: how much can we rely on published data on potential drug targets? Nat. Rev. Drug Discov. 10, 712 (2011).

    Article 

    Google Scholar 

  • Begley, C. G. & Ellis, L. M. Raise standards for preclinical cancer research. Nature 483, 531–533 (2012).

    Article 

    Google Scholar 

  • Gelman, A. & Loken, E. The Garden of Forking Paths: Why Multiple Comparisons Can be a Problem, Even When There is no ‘Fishing Expedition’ or ‘p-Hacking’ and the Research Hypothesis was Posited Ahead of Time Vol. 348, 1–17 (Department of Statistics, 2013).

  • Baker, M. Reproducibility crisis. Nature 533, 353–66 (2016).

    Google Scholar 

  • Karagiorgi, G., Kasieczka, G., Kravitz, S., Nachman, B. & Shih, D. Machine learning in the search for new fundamental physics. Nat. Rev. Phys. 4, 399–412 (2022).

    Article 

    Google Scholar 

  • Dara, S., Dhamercherla, S., Jadav, S. S., Babu, C. H. M. & Ahsan, M. J. Machine learning in drug discovery: a review. Artif. Intell. Rev. 55, 1947–1999 (2022).

    Article 

    Google Scholar 

  • Mater, A. C. & Coote, M. L. Deep learning in chemistry. J. Chem. Inform. Model. 59, 2545–2559 (2019).

    Article 

    Google Scholar 

  • Carleo, G. et al. Machine learning and the physical sciences. Rev. Modern Phys. 91, 045002 (2019).

    Article 

    Google Scholar 

  • Rajkomar, A., Dean, J. & Kohane, I. Machine learning in medicine. New England J. Med. 380, 1347–1358 (2019).

    Article 

    Google Scholar 

  • Grimmer, J., Roberts, M. E. & Stewart, B. M. Machine learning for social science: an agnostic approach. Ann. Rev. Pol. Sci. 24, 395–419 (2021).

    Article 

    Google Scholar 

  • Jumper, J. et al. Highly accurate protein structure prediction with alphafold. Nature 596, 583–589 (2021).

    Article 

    Google Scholar 

  • Hutson, M. Artificial intelligence faces reproducibility crisis. Science 359, 725–726 (2018).

    Article 

    Google Scholar 

  • Gundersen, O.E., Coakley, K., Kirkpatrick, C. & Gil, Y. Sources of irreproducibility in machine learning: a review. Preprint at (2022).

  • Sculley, D., Snoek, J., Wiltschko, A. & Rahimi, A. Winner’s Curse? On Pace, Progress, and Empirical Rigor (ICLR, 2018).

  • Armstrong, T. G., Moffat, A., Webber, W. & Zobel, J. Improvements that don’t add up: ad-hoc retrieval results since 1998. In Proc. 18th ACM Conference on Information and Knowledge Management 601–610 (ACM, 2009).

  • Kapoor, S. & Narayanan, A. Leakage and the reproducibility crisis in machine-learning-based science. Patterns, 4, 100804 (2023).

  • Kapoor, S. et al. Reforms: reporting standards for machine learning based science. Preprint at (2023).

  • DeMasi, O., Kording, K. & Recht, B. Meaningless comparisons lead to false optimism in medical machine learning. PloS ONE 12, e0184604 (2017).

    Article 

    Google Scholar 

  • Roberts, M. et al. Common pitfalls and recommendations for using machine learning to detect and prognosticate for COVID-19 using chest radiographs and CT scans. Nat. Mach. Intell. 3, 199–217 (2021).

    Article 

    Google Scholar 

  • Wynants, L. et al. Prediction models for diagnosis and prognosis of COVID-19: systematic review and critical appraisal. BMJ 369, m1328 (2020).

  • Whalen, S., Schreiber, J., Noble, W. S. & Pollard, K. S. Navigating the pitfalls of applying machine learning in genomics. Nat. Rev. Genet. 23, 169–181 (2022).

    Article 

    Google Scholar 

  • Artrith, N. et al. Best practices in machine learning for chemistry. Nat. Chem. 13, 505–508 (2021).

    Article 

    Google Scholar 

  • Thuerey, N. et al. Physics-based deep learning. Preprint at (2021).

  • Brunton, S. L. & Kutz, J. N. Promising directions of machine learning for partial differential equations. Nat. Comput. Sci. 4, 483–494 (2024).

  • Vinuesa, R. & Brunton, S. L. Enhancing computational fluid dynamics with machine learning. Nat. Comput. Sci. 2, 358–366 (2022).

    Article 

    Google Scholar 

  • Karniadakis, G. E. et al. Physics-informed machine learning. Nat. Rev. Phys. 3, 422–440 (2021).

    Article 

    Google Scholar 

  • Cuomo, S. et al. Scientific machine learning through physics–informed neural networks: where we are and what’s next. J. Sci. Comput. 92, 88 (2022).

    Article 
    MathSciNet 

    Google Scholar 

  • Duraisamy, K., Iaccarino, G. & Xiao, H. Turbulence modeling in the age of data. Ann. Rev. Fluid Mech. 51, 357–377 (2019).

    Article 
    MathSciNet 

    Google Scholar 

  • Durran, D. R. Numerical Methods for Wave Equations in Geophysical Fluid Dynamics Vol. 32 (Springer, 2013).

  • LeVeque, R. J. Numerical Methods for Conservation Laws Vol. 214 (Birkhäuser, 1992).

  • Mishra, S. A machine learning framework for data driven acceleration of computations of differential equations. Math. Eng. (2018).

  • Kochkov, D. et al. Machine learning–accelerated computational fluid dynamics. Proc. Natl Acad. Sci. USA 118, e2101784118 (2021).

    Article 
    MathSciNet 

    Google Scholar 

  • Kadapa, C. Machine learning for computational science and engineering—a brief introduction and some critical questions. Preprint at (2021).

  • Ross, A., Li, Z., Perezhogin, P., Fernandez-Granda, C. & Zanna, L. Benchmarking of machine learning ocean subgrid parameterizations in an idealized model. J. Adv. Model. Earth Syst. 15, e2022MS003258 (2023).

    Article 

    Google Scholar 

  • Lippe, P., Veeling, B., Perdikaris, P., Turner, R. & Brandstetter, J. PDE-refiner: achieving accurate long rollouts with neural PDE solvers. In 37th Conference on Neural Information Processing Systems (NeurIPS 2023).

  • Vlachas, P. R. et al. Backpropagation algorithms and reservoir computing in recurrent neural networks for the forecasting of complex spatiotemporal dynamics. Neural Netw. 126, 191–217 (2020).

    Article 

    Google Scholar 

  • Raissi, M., Perdikaris, P. & Karniadakis, G. E. Physics-informed neural networks: a deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations. J. Comput. Phys. 378, 686–707 (2019).

    Article 
    MathSciNet 

    Google Scholar 

  • Grossmann, T. G., Komorowska, U. J., Latz, J. & Schönlieb, C.-B. Can physics-informed neural networks beat the finite element method? IMA J. Appl. Math. 89, 143–174, (2024).

  • de la Mata, F. F., Gijón, A., Molina-Solana, M. & Gómez-Romero, J. Physics-informed neural networks for data-driven simulation: advantages, limitations, and opportunities. Phys. A 610, 128415 (2023).

    Article 
    MathSciNet 

    Google Scholar 

  • Chuang, P.-Y. & Barba, L. A. Experience report of physics-informed neural networks in fluid simulations: pitfalls and frustration. Preprint at (2022).

  • Chuang, P.-Y. & Barba, L. A. Predictive limitations of physics-informed neural networks in vortex shedding. Preprint at (2023).

  • Wang, S., Yu, X. & Perdikaris, P. When and why pinns fail to train: a neural tangent kernel perspective. J. Comput. Phys. 449, 110768 (2022).

    Article 
    MathSciNet 

    Google Scholar 

  • Krishnapriyan, A., Gholami, A., Zhe, S., Kirby, R. & Mahoney, M. W. Characterizing possible failure modes in physics-informed neural networks. In 35th Conference on Neural Information Processing Systems Vol. 34, 26548–26560 (NeurIPS 2021).

  • Basir, S. & Senocak, I. Critical investigation of failure modes in physics-informed neural networks. In AiAA SCITECH 2022 Forum 2353 (ARC, 2022).

  • Karnakov, P., Litvinov, S. & Koumoutsakos, P. Solving inverse problems in physics by optimizing a discrete loss: fast and accurate learning without neural networks. Proc. Natl Acad. Sci. Nexus 3, pgae005 (2024).

  • Gundersen, O. E. The fundamental principles of reproducibility. Phil. Trans. R. Soc. A 379, 20200210 (2021).

    Article 

    Google Scholar 

  • Aromataris, E. & Pearson, A. The systematic review: an overview. Am. J. Nursing 114, 53–58 (2014).

    Article 

    Google Scholar 

  • Magiera, J., Ray, D., Hesthaven, J. S. & Rohde, C. Constraint-aware neural networks for riemann problems. J. Comput. Phys. 409, 109345 (2020).

    Article 
    MathSciNet 

    Google Scholar 

  • Bezgin, D. A., Schmidt, S. J. & Adams, N. A. A data-driven physics-informed finite-volume scheme for nonclassical undercompressive shocks. J. Comput. Phys. 437, 110324 (2021).

    Article 
    MathSciNet 

    Google Scholar 

  • Dresdner, G. et al. Learning to correct spectral methods for simulating turbulent flows. Transact. Mach. Learn Res. (2023).

  • Toshev, A. P., Galletti, G., Brandstetter, J., Adami, S. & Adams, N. A. E(3) equivariant graph neural networks for particle-based fluid mechanics. In ICLR 2023 Workshop on Physics for Machine Learning (2023).

  • Discacciati, N., Hesthaven, J. S. & Ray, D. Controlling oscillations in high-order discontinuous galerkin schemes using artificial viscosity tuned by neural networks. J. Comput. Phys. 409, 109304 (2020).

    Article 
    MathSciNet 

    Google Scholar 

  • Dong, S. & Yang, J. On computing the hyperparameter of extreme learning machines: algorithm and application to computational pdes, and comparison with classical and high-order finite elements. J. Comput. Phys. 463, 111290 (2022).

    Article 
    MathSciNet 

    Google Scholar 

  • Bar-Sinai, Y., Hoyer, S., Hickey, J. & Brenner, M. P. Learning data-driven discretizations for partial differential equations. Proc. Natl Acad. Sci. USA 116, 15344–15349 (2019).

    Article 
    MathSciNet 

    Google Scholar 

  • Li, Z. et al. Fourier neural operator for parametric partial differential equations. In Proc. of ICLR 2021 (2021).

  • Li, Z. et al. Physics-informed neural operator for learning partial differential equations. ACM/JMS J. Data Sci. (2024).

  • Richards, G. C. & Onakpoya, I. J. in Catalogue of Bias (Catalogue of Bias, 2019); https://catalogofbias.org/biases/reporting-biases/

  • Thornton, A. & Lee, P. Publication bias in meta-analysis: its causes and consequences. J. Clin. Epidemiol. 53, 207–216 (2000).

    Article 

    Google Scholar 

  • Boutron, I. & Ravaud, P. Misrepresentation and distortion of research in biomedical literature. Proc. Natl Acad. Sci. USA 115, 2613–2619 (2018).

    Article 

    Google Scholar 

  • Thomas, E. T. & Heneghan, C. Catalogue of bias: selective outcome reporting bias. BMJ Evid.-Based Med. 27, 370–372 (2022).

    Article 

    Google Scholar 

  • Head, M. L., Holman, L., Lanfear, R., Kahn, A. T. & Jennions, M. D. The extent and consequences of p-hacking in science. PLoS Biol. 13, e1002106 (2015).

    Article 

    Google Scholar 

  • Saini, P. et al. Selective reporting bias of harm outcomes within studies: findings from a cohort of systematic reviews. BMJ 349, g6501 (2014).

  • Altman, D. G., Moher, D. & Schulz, K. F. Harms of outcome switching in reports of randomised trials: consort perspective. BMJ 356, j396 (2017).

  • Erasmus, A., Holman, B. & Ioannidis, J. P. A. Data-dredging bias. BMJ Evid.-Based Med. 27, 209–211 (2022).

    Article 

    Google Scholar 

  • De Vries, Y. A. et al. The cumulative effect of reporting and citation biases on the apparent efficacy of treatments: the case of depression. Psychol. Med. 48, 2453–2455 (2018).

    Article 

    Google Scholar 

  • Fanelli, D. ‘Positive’ results increase down the hierarchy of the sciences. PloS ONE 5, e10068 (2010).

    Article 

    Google Scholar 

  • Wang, S., Wang, H. & Perdikaris, P. Learning the solution operator of parametric partial differential equations with physics-informed deeponets. Sci. Adv. 7, eabi8605 (2021).

    Article 

    Google Scholar 

  • Um, K., Brand, R., Fei, Y. R., Holl, P. & Thuerey, N. Solver-in-the-loop: learning from differentiable physics to interact with iterative PDE-solvers. In 34th Conference on Neural Information Processing Systems 33, 6111–6122 (NeurIPS, 2020).

  • Brandstetter, J., Worrall, D. & Welling, M. Message passing neural PDE solvers. In Proc. of ICLR 2022 (2022).

  • Dong, S. & Li, Z. Local extreme learning machines and domain decomposition for solving linear and nonlinear partial differential equations. Comput. Methods Appl. Mech. Eng. 387, 114129 (2021).

    Article 
    MathSciNet 

    Google Scholar 

  • Shang, Y., Wang, F. & Sun, J. Deep Petrov–Galerkin method for solving partial differential equations. Preperint at (2022).

  • Li, Z., Huang, D. Z., Liu, B. & Anandkumar, A. Fourier neural operator with learned deformations for PDEs on general geometries. J. Mach. Learn. Res. (2023).

  • Zhuang, J., Kochkov, D., Bar-Sinai, Y., Brenner, M. P. & Hoyer, S. Learned discretizations for passive scalar advection in a two-dimensional turbulent flow. Phys. Rev. Fluids 6, 064605 (2021).

    Article 

    Google Scholar 

  • Kube, R., Churchill, R.M. & Sturdevant, B. Machine learning accelerated particle-in-cell plasma simulations. In Proc. of Fourth Workshop on Machine Learning and the Physical Sciences (NeurIPS 2021) (2021).

  • Stevens, B. & Colonius, T. FiniteNet: a fully convolutional LSTM network architecture for time-dependent partial differential equations. Preprint at (2020).

  • Alguacil, A., Bauerheim, M., Jacob, M. C. & Moreau, S. Predicting the propagation of acoustic waves using deep convolutional neural networks. J. Sound Vib. 512, 116285 (2021).

    Article 

    Google Scholar 

  • Alguacil, A., Bauerheim, M., Jacob, M. C. & Moreau, S. Deep learning surrogate for the temporal propagation and scattering of acoustic waves. AIAA J. 60, 5890–5906 (2022).

    Article 

    Google Scholar 

  • Bezgin, D. A., Schmidt, S. J. & Adams, N. A. WENO3-NN: a maximum-order three-point data-driven weighted essentially non-oscillatory scheme. J. Comput. Phys. 452, 110920 (2022).

    Article 
    MathSciNet 

    Google Scholar 

  • Xiao, X., Zhou, Y., Wang, H. & Yang, X. A novel cnn-based poisson solver for fluid simulation. IEEE Trans. Vis. Comput. Graphics 26, 1454–1465 (2018).

    Article 

    Google Scholar 

  • Sanchez-Gonzalez, A. et al. Learning to simulate complex physics with graph networks. In International Conference on Machine Learning 8459–8468 (PMLR, 2020).

  • Klimesch, J., Holl, P. & Thuerey, N. Simulating liquids with graph networks. Preprint at (2022).

  • Wang, S., Wang, H. & Perdikaris, P. Improved architectures and training algorithms for deep operator networks. J. Sci. Comput. 92, 35 (2022).

    Article 
    MathSciNet 

    Google Scholar 

  • Gupta, J. K. & Brandstetter, J. Towards multi-spatiotemporal-scale generalized PDE modeling. Transact. Mach. Lean. Res. (2023).

  • McGreivy, N. & Hakim, A. Invariant preservation in machine learned PDE solvers via error correction. Preprint at (2023).

  • Stachenfeld, K. et al. Learned simulators for turbulence. In Proc. of International Conference on Learning Representations (ICLR) 2022 (2022).

  • Cheng, L., Illarramendi, E. A., Bogopolsky, G., Bauerheim, M. & Cuenot, B. Using neural networks to solve the 2D Poisson equation for electric field computation in plasma fluid simulations. Preprint at (2021).

  • Simmons, J. P., Nelson, L. D. & Simonsohn, U. False-positive psychology: undisclosed flexibility in data collection and analysis allows presenting anything as significant. Psychol. Sci. 22, 1359–1366 (2011).

    Article 

    Google Scholar 

  • Wicherts, J. M. et al. Degrees of freedom in planning, running, analyzing, and reporting psychological studies: a checklist to avoid p-hacking. Front. Psychol. 7, 1832 (2016).

  • Serra-Garcia, M. & Gneezy, U. Nonreplicable publications are cited more than replicable ones. Sci. Adv. 7, eabd1705 (2021).

    Article 

    Google Scholar 

  • Borji, A. Negative results in computer vision: a perspective. Image Vis. Comput. 69, 1–8 (2018).

    Article 

    Google Scholar 

  • Smith, J. J., Amershi, S., Barocas, S., Wallach, H. & Wortman Vaughan, J. Real ML: recognizing, exploring, and articulating limitations of machine learning research. In 2022 ACM Conference on Fairness, Accountability, and Transparency 587–597 (ACM, 2022).

  • Gundersen, O. E. The case against registered reports. AI Magazine 42, 88–92 (2021).

    Article 

    Google Scholar 

  • Schooler, J. W. Metascience could rescue the ‘replication crisis’. Nature 515, 9–9 (2014).

    Article 

    Google Scholar 

  • Ray, D. & Hesthaven, J. S. An artificial neural network as a troubled-cell indicator. J. Comput. Phys. 367, 166–191 (2018).

    Article 
    MathSciNet 

    Google Scholar 

  • Wang, S. & Perdikaris, P. Long-time integration of parametric evolution equations with physics-informed deeponets. J. Comput. Phys. 475, 111855 (2023).

    Article 
    MathSciNet 

    Google Scholar 

  • Ovadia, O., Kahana, A., Turkel, E. & Dekel, S. Beyond the Courant–Friedrichs–Lewy condition: numerical methods for the wave problem using deep learning. J. Comput. Phys. 442, 110493 (2021).

    Article 
    MathSciNet 

    Google Scholar 

  • Li, Z. et al. Learning dissipative dynamics in chaotic systems. In Proc. of 36th Conference on Neural Information Processing Systems (NeurIPS) (2022).

  • Ni, N. & Dong, S. Numerical computation of partial differential equations by hidden-layer concatenated extreme learning machine. J. Sci. Comput. 95, 35 (2023).

    Article 
    MathSciNet 

    Google Scholar 

  • Dong, S. & Li, Z. A modified batch intrinsic plasticity method for pre-training the random coefficients of extreme learning machines. J. Comput. Phys. 445, 110585 (2021).

    Article 
    MathSciNet 

    Google Scholar 

  • Mueller, M., Greif, R., Jenko, F. & Thuerey, N. Leveraging thestochastic predictions of Bayesian neural networks for fluid simulations. In Machine Learning and the Physical Sciences workshop, NeurIPS 2022 (2022).

  • Wang, S., Bhouri, M. A. & Perdikaris, P. Fast PDE-constrained optimization via self-supervised operator learning. Preprint at (2021).

  • Schwander, L., Ray, D. & Hesthaven, J. S. Controlling oscillations in spectral methods by local artificial viscosity governed by neural networks. J. Comput. Phys. 431, 110144 (2021).

    Article 
    MathSciNet 

    Google Scholar 

  • Donon, B. et al. Deep statistical solvers. In 34th Conference on Neural Information Processing Systems Vol. 33, 7910–7921 (NeurIPS 2020).

  • Wan, Z. Y., Zepeda-Núñez, L., Boral, A. & Sha, F. Evolve smoothly, fit consistently: learning smooth latent dynamics for advection-dominated systems. In Proc. of International Conference on Learning Representaton (ICLR) (2023).

  • Di Leoni, P. C., Lu, L., Meneveau, C., Karniadakis, G. E. & Zaki, T. A. Neural operator prediction of linear instability waves in high-speed boundary layers. J. Comput. Phys. 474, 111793 (2023).

    Article 
    MathSciNet 

    Google Scholar 

  • Kovachki, N. et al. Neural operator: learning maps between function spaces with applications to PDEs. J. of Mach. Learn. Res. (2023).

  • Holl, P., Koltun, V., Um, K. & Thuerey, N. phiflow: a differentiable PDE solving framework for deep learning via physical simulations. In Workshop on Differentiable Vision, Graphics, and Physics in Machine Learning at NeurIPS (NeurIPS, 2020).

  • Nemmen, R., Duarte, R. & Navarro, J. P. The first AI simulation of a black hole. Proc. Intl Astron. Union 15, 329–333 (2020).

    Article 

    Google Scholar 

  • Wandel, N., Weinmann, M. & Klein, R. Learning incompressible fluid dynamics from scratch–towards fast, differentiable fluid models that generalize. In Proc. of International Conference on Learning Representations (ICLR) 2021 (2021).

  • Haridas, A., Vadlamani, N. R. & Minamoto, Y. Deep neural networks to correct sub-precision errors in CFD. Appl. Energy Combustion Sci. 12, 100081 (2022).

    Article 

    Google Scholar 

  • Guennebaud, G. et al. Eigen v 3 (Eigen, 2010); http://eigen.tuxfamily.org

  • Falgout, R. D. & Yang, U. M. hypre: A library of high performance preconditioners. In International Conference on Computational Science 632–641 (Springer, 2002).

  • Mayr, M., Wall, W. A. & Gee, M. W. Adaptive time stepping for fluid–structure interaction solvers. Finite Elements in Anal. Design 141, 55–69 (2018).

    Article 
    MathSciNet 

    Google Scholar 

  • Reynolds, D. R., Woodward, C. S., Gardner, D. J. & Hindmarsh, A. C. Arkode: a library of high order implicit/explicit methods for multi-rate problems. In SIAM Conference on Parallel Processing for Scientific Computing (2014).

  • Gottlieb, S., Shu, C.-W. & Tadmor, E. Strong stability-preserving high-order time discretization methods. SIAM Rev. 43, 89–112 (2001).

    Article 
    MathSciNet 

    Google Scholar 

  • Hakim, A., Hammett, G., Shi, E. & Mandell, N. Discontinuous galerkin schemes for a class of hamiltonian evolution equations with applications to plasma fluid and kinetic problems. Preprint at (2019).

  • Cockburn, B. & Shu, C.-W. Tvb runge-kutta local projection discontinuous galerkin finite element method for conservation laws. ii. General framework. Math. Comput. 52, 411–435 (1989).

    MathSciNet 

    Google Scholar 

  • Qin, T., Beatson, A., Oktay, D., McGreivy, N. & Adams, R. P. Meta-PDE: learning to solve pdes quickly without a mesh. Preprint at (2022).

  • McGreivy, N. & Hakim, A. Data Created During Random Sampling From and Systematic Review of ML-for-PDE Solving Research (OSF, 2024); https://doi.org/10.17605/OSF.IO/GQ5B3

  • McGreivy, N. nickmcgreivy/WeakBaselinesMLPDE: First release (v1.0) (Zenodo, 2024); https://doi.org/10.5281/zenodo.12682908

  • McGreivy, N. & Hakim, A. Weak baselines and reporting biases lead to overoptimism in machine learning for fluid-related partial differential equations. Code Ocean (2024).

  • McGreivy, N. & Hakim, A. (GPU Code) Weak baselines and reporting biases lead to overoptimism in machine learning for fluid-related partial differential equations. Code Ocean (2024).

  • Lu, L., Jin, P., Pang, G., Zhang, Z. & Karniadakis, G. E. Learning nonlinear operators via deeponet based on the universal approximation theorem of operators. Nat. Mach, Intell. 3, 218–229 (2021).

    Article 

    Google Scholar 

  • Tompson, J., Schlachter, K., Sprechmann, P. & Perlin, K. Accelerating eulerian fluid simulation with convolutional networks. In International Conference on Machine Learning, 3424–3433 (PMLR, 2017).

  • Pfaff, T., Fortunato, M., Sanchez-Gonzalez, A. & Battaglia, P. W. Learning mesh-based simulation with graph networks. In Proc. of International Conference on Learning Representations (ICLR) 2021 (2021).

  • Kim, B. et al. Deep fluids: a generative network for parameterized fluid simulations. In Computer Graphics Forum Vol. 38, 59–70 (Wiley, 2019).

  • Cai, S., Wang, Z., Lu, L., Zaki, T. A. & Karniadakis, G. E. DeepM&Mnet: inferring the electroconvection multiphysics fields based on operator approximation by neural networks. J. Comput. Phys. 436, 110296 (2021).

    Article 
    MathSciNet 

    Google Scholar 

  • De Avila Belbute-Peres, F., Economon, T. & Kolter, Z. Combining differentiable PDE solvers and graph neural networks for fluid flow prediction. In International Conference on Machine Learning 2402–2411 (PMLR, 2020).

  • Yang, C., Yang, X. & Xiao, X. Data-driven projection method in fluid simulation. Comput. Animation Virtual Worlds 27, 415–424 (2016).

    Article 

    Google Scholar 

  • Hsieh, J.-T., Zhao, S., Eismann, S., Mirabella, L. & Ermon, S. Learning neural PDE solvers with convergence guarantees. In Proc. of International Conference on Learning Representations (ICLR) 2021 (2021).

  • Mao, Z., Lu, L., Marxen, O., Zaki, T. A. & Karniadakis, G. E. DeepM&Mnet for hypersonics: predicting the coupled flow and finite-rate chemistry behind a normal shock using neural-network approximation of operators. J. Comput. Phys. 447, 110698 (2021).

    Article 
    MathSciNet 

    Google Scholar 

  • Greenfeld, D., Galun, M., Basri, R., Yavneh, I. & Kimmel, R. Learning to optimize multigrid PDE solvers. In International Conference on Machine Learning 2415–2423 (PMLR, 2019).

  • Ray, D. & Hesthaven, J. S. Detecting troubled-cells on two-dimensional unstructured grids using a neural network. J. Comput. Phys. 397, 108845 (2019).

    Article 
    MathSciNet 

    Google Scholar 

  • Shan, T. et al. Study on a fast solver for poisson’s equation based on deep learning technique. IEEE Trans. Antennas Propagation 68, 6725–6733 (2020).

    Article 

    Google Scholar 

  • Luz, I., Galun, M., Maron, H., Basri, R. & Yavneh, I. Learning algebraic multigrid using graph neural networks. In International Conference on Machine Learning 6489–6499 (PMLR, 2020).

  • Pathak, J. et al. Using machine learning to augment coarse-grid computational fluid dynamics simulations. Preprint at (2020).

  • Di Leoni, P. C., Lu, L., Meneveau, C., Karniadakis, G. & Zaki, T. A. Neural operator prediction of linear instability waves in high-speed boundary layers. J. Comput. Phys. (2023).

  • Stevens, B. & Colonius, T. Enhancement of shock-capturing methods via machine learning. Theor. Comput. Fluid Dyn. 34, 483–496 (2020).

    Article 
    MathSciNet 

    Google Scholar 

  • Illarramendi, E. A. et al. Towards an hybrid computational strategy based on deep learning for incompressible flows. In AIAA Aviation 2020 Forum 3058 (AIAA, 2020).

  • Han, X., Gao, H., Pfaff, T., Wang, J.-X. & Liu, L.-P. Predicting physics in mesh-reduced space with temporal attention. In Proc. of the International Conference on Learning Responses (ICLR) 2022 (2022).

  • Özbay, A. G. et al. Poisson CNN: convolutional neural networks for the solution of the poisson equation on a cartesian mesh. Data-Centric Eng. 2, e6 (2021).

    Article 

    Google Scholar 

  • Li, Z. & Farimani, A. B. Graph neural network-accelerated Lagrangian fluid simulation. Comput. Graphics 103, 201–211 (2022).

    Article 

    Google Scholar 

  • Peng, W., Yuan, Z. & Wang, J. Attention-enhanced neural network models for turbulence simulation. Phys. Fluids, 34, 025111 (2022).

  • Chen, L.-W., Cakal, B. A., Hu, X. & Thuerey, N. Numerical investigation of minimum drag profiles in laminar flow using deep learning surrogates. J. Fluid Mech. 919, A34 (2021).

    Article 
    MathSciNet 

    Google Scholar 

  • Wandel, N., Weinmann, M. & Klein, R. Teaching the incompressible navier–stokes equations to fast neural surrogate models in three dimensions. Phys. Fluids, 33, 047117 (2021).

  • List, B., Chen, L.-W. & Thuerey, N. Learned turbulence modelling with differentiable fluid solvers: physics-based loss functions and optimisation horizons. J. Fluid Mech. 949, A25 (2022).

    Article 
    MathSciNet 

    Google Scholar 

  • Wen, X., Don, W. S., Gao, Z. & Hesthaven, J. S. An edge detector based on artificial neural network with application to hybrid compact-weno finite difference scheme. J. Sci. Comput. 83, 1–21 (2020).

    Article 
    MathSciNet 

    Google Scholar 

  • de Lara, F. M. & Ferrer, E. Accelerating high order discontinuous galerkin solvers using neural networks: 1D Burgers’ equation. Comput. Fluids 235, 105274 (2022).

    Article 
    MathSciNet 

    Google Scholar 

  • Zhao, Q., Lindell, D. B. & Wetzstein, G. Learning to solve PDE-constrained inverse problems with graph networks. In Proc. of the 39th International Conference on Machine Learning (eds Chaudhuri, K et al.) 162, 26895–26910 (2022).

  • Illarramendi, E. A., Bauerheim, M. & Cuenot, B. Performance and accuracy assessments of an incompressible fluid solver coupled with a deep convolutional neural network. Data-Centric Engineering 3, e2 (2022).

    Article 

    Google Scholar 

  • Holloway, I., Wood, A. & Alekseenko, A. Acceleration of boltzmann collision integral calculation using machine learning. Mathematics 9, 1384 (2021).

    Article 

    Google Scholar 

  • Azulay, Y. & Treister, E. Multigrid-augmented deep learning preconditioners for the helmholtz equation. SIAM J. Sci. Comput. 45, S127–S151 (2022).

    Article 
    MathSciNet 

    Google Scholar 

  • Wu, T., Maruyama, T. & Leskovec, J. Learning to accelerate partial differential equations via latent global evolution. In 36th Conference on Neural Information Processing Systems Vol. 35, 2240–2253 (NeurIPS 2022).

  • Liu, X.-Y., Sun, H., Zhu, M., Lu, L. & Wang, J.-X. Predicting parametric spatiotemporal dynamics by multi-resolution PDE structure-preserved deep learning. Preprint at (2022).

  • Zhang, E. et al. A hybrid iterative numerical transferable solver (hints) for pdes based on deep operator network and relaxation methods. Preprint at (2022).

  • Duarte, R., Nemmen, R. & Navarro, J. P. Black hole weather forecasting with deep learning: a pilot study. Monthly Notices R. Astron. Soc. 512, 5848–5861 (2022).

    Article 

    Google Scholar 

  • Huang, X. et al. LordNet: an efficient neural network for learning to solve parametric partial differential equations without simulated data. Neural Netw. (2024).

  • Ranade, R., Hill, C., He, H., Maleki, A. & Pathak, J. A latent space solver for PDE generalization. In ICLR 2021 SimDL Workshop (2021).

  • Chen, R., Jin, X. & Li, H. A machine learning based solver for pressure poisson equations. Theor. Appl. Mech. Lett. 12, 100362 (2022).

    Article 

    Google Scholar 

  • Ranade, R. et al. A composable autoencoder-based iterative algorithm for accelerating numerical simulations. Preprint at (2021).

  • Peng, W., Yuan, Z., Li, Z. & Wang, J. Linear attention coupled fourier neural operator for simulation of three-dimensional turbulence. Phys. Fluids, 35, 015106 (2023).

  • de Lara, F. M. & Ferrer, E. Accelerating high order discontinuous galerkin solvers using neural networks: 3D compressible Navier–Stokes equations. J. Comput. Phys. 489, 112253, (2023).

  • Ranade, R., Hill, C., Ghule, L. & Pathak, J. A composable machine-learning approach for steady-state simulations on high-resolution grids. In 36th Conference on Neural Information Processing Systems Vol. 35, 17386–17401 (NeurIPS, 2022).

  • Fang, D. & Tan, J. Immersed boundary-physics informed machine learning approach for fluid–solid coupling. Ocean Eng. 263, 112360 (2022).

    Article 

    Google Scholar 

  • Shukla, K. et al. Deep neural operators can serve as accurate surrogates for shape optimization. Eng. Appl. Artif. Intell. (2024).

  • Zhang, T., Innanen, K. & Trad, D. Learning the elastic wave equation with fourier neural operators. Geoconvention 2022, 1–5 (2022).

    Google Scholar 

  • Bezgin, D. A., Buhendwa, A. B. & Adams, N. A. A fully-differentiable compressible high-order computational fluid dynamics solver. Preprint at (2021).

  • Yang, Y., Gao, A. F., Azizzadenesheli, K., Clayton, R. W. & Ross, Z. E. Rapid seismic waveform modeling and inversion with neural operators. IEEE Trans. Geosci. Remote Sensing 61, 1–12 (2023).

    Google Scholar 

  • Tang, J., Azevedo, V. C., Cordonnier, G. & Solenthaler, B. Neural green’s function for laplacian systems. Comput. Graphics 107, 186–196 (2022).

    Article 

    Google Scholar 

  • Nastorg, M. et al. DS-GPS: a deep statistical graph poisson solver (for faster CFD simulations). In Machine Learning and the Physical Sciences workshop, NeurIPS 2022 (2022).

  • Gopakumar, V. et al. Fourier neural operator for plasma modelling. Preprint at (2023).

  • Shit, S. et al. Semi-implicit neural solver for time-dependent partial differential equations. Preprint at (2021).

  • Su, X., Walters, R., Aslangil, D. & Yu, R. Forecasting variable-density 3D turbulent flow. In ICLR 2021 SimDL Workshop (ICLR, 2021).

  • Jeon, J., Lee, J., Eivazi, H., Vinuesa, R. & Kim, S. J. Residual-based physics-informed transfer learning: a hybrid method for accelerating long-term CFD simulations via deep learning. Int. J. Heat Mass Transf. (2024).

  • Dai, Y., An, Y. & Li, Z. Fournetflows: an efficient model for steady airfoil flows prediction. Preprint at (2022).

  • Sun, J., Dong, S. & Wang, F. Local randomized neural networks with discontinuous Galerkin methods for partial differential equations. J. Comput. App. Math. (2024).

  • Shao, T., Shan, T., Li, M., Yang, F. & Xu, S. A poisson’s equation solver based on neural network precondtioned CG method. In 2022 International Applied Computational Electromagnetics Society Symposium (ACES-China) 1–2 (IEEE, 2022).

  • link

    Leave a Reply

    Your email address will not be published. Required fields are marked *