Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
169 tokens/sec
GPT-4o
7 tokens/sec
Gemini 2.5 Pro Pro
45 tokens/sec
o3 Pro
4 tokens/sec
GPT-4.1 Pro
38 tokens/sec
DeepSeek R1 via Azure Pro
28 tokens/sec
2000 character limit reached

A Unified Framework for the Error Analysis of Physics-Informed Neural Networks (2311.00529v2)

Published 1 Nov 2023 in math.NA and cs.NA

Abstract: We prove a priori and a posteriori error estimates for physics-informed neural networks (PINNs) for linear PDEs. We analyze elliptic equations in primal and mixed form, elasticity, parabolic, hyperbolic and Stokes equations; and a PDE constrained optimization problem. For the analysis, we propose an abstract framework in the common language of bilinear forms, and we show that coercivity and continuity lead to error estimates. The obtained estimates are sharp and reveal that the $L2$ penalty approach for initial and boundary conditions in the PINN formulation weakens the norm of the error decay. Finally, utilizing recent advances in PINN optimization, we present numerical examples that illustrate the ability of the method to achieve accurate solutions.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (52)
  1. On the existence and regularity of the solution of Stokes problem in arbitrary dimension. Proc. Japan Acad. Ser. A Math. Sci, 67 (5):171–175, 1991.
  2. Least squares methods for elliptic systems. Mathematics of Computation, 44(169):53–70, 1985.
  3. JAX: composable transformations of Python+NumPy programs, 2018.
  4. Rayleigh-Ritz-Galerkin methods for Dirichlet’s problem using subspaces without boundary conditions. Communications on Pure and Applied Mathematics, 23(4):653–675, 1970.
  5. The Mathematical Theory of Finite Element Methods. Springer, 2008.
  6. Artificial intelligence velocimetry and microaneurysm-on-a-chip for three-dimensional analysis of blood flow in physiology and disease. Proceedings of the National Academy of Sciences, 118(13):e2100697118, 2021.
  7. Physics-informed neural networks (PINNs) for fluid mechanics: A review. Acta Mechanica Sinica, 37(12):1727–1738, 2021.
  8. On a theorem by Lions and Peetre about interpolation between a Banach space and its dual. Houston J. Math, 24(2):325–344, 1998.
  9. Error estimates for physics-informed neural networks approximating the Navier–Stokes equations. IMA Journal of Numerical Analysis, page drac085, 01 2023.
  10. Error analysis for physics-informed neural networks (PINNs) approximating Kolmogorov PDEs. Advances in Computational Mathematics, 48(6):79, 2022.
  11. Neural-network-based approximations for solving partial differential equations. Communications in Numerical Methods in Engineering, 10(3):195–201, 1994.
  12. Theory and Practice of Finite Elements, volume 159. Springer, 2004.
  13. Lawrence C Evans. Partial Differential Equations, volume 19. American Mathematical Society, 2022.
  14. J-L Guermond. The LBB condition in fractional Sobolev spaces and applications. IMA Journal of Numerical Analysis, 29(3):790–805, 2009.
  15. Approximation rates for neural networks with encodable weights in smoothness spaces. Neural Networks, 134:107–130, 2021.
  16. A physics-informed deep learning framework for inversion and surrogate modeling in solid mechanics. Computer Methods in Applied Mechanics and Engineering, 379:113741, 2021.
  17. Solving high-dimensional partial differential equations using deep learning. Proceedings of the National Academy of Sciences, 115(34):8505–8510, 2018.
  18. Deep learning-based numerical methods for high-dimensional parabolic partial differential equations and backward stochastic differential equations. Communications in Mathematics and Statistics, 5(4):349–380, 2017.
  19. Nvidia simnet™: An ai-accelerated multi-physics simulation framework. In International conference on computational science, pages 447–461. Springer, 2021.
  20. Deep-neural-network solution of the electronic Schrödinger equation. Nature Chemistry, 12(10):891–897, 2020.
  21. Solving Poisson problems in polygonal domains with singularity enriched physics informed neural networks. arXiv preprint arXiv:2308.16429, 2023.
  22. Analysis in Banach spaces, volume 12. Springer, 2016.
  23. Adaptive activation functions accelerate convergence in deep and physics-informed neural networks. Journal of Computational Physics, 404:109136, 2020.
  24. NSFnets (Navier-Stokes flow nets): Physics-informed neural networks for the incompressible Navier-Stokes equations. Journal of Computational Physics, 426:109951, 2021.
  25. Physics-informed machine learning. Nature Reviews Physics, 3(6):422–440, 2021.
  26. Characterizing possible failure modes in physics-informed neural networks. Advances in Neural Information Processing Systems, 34:26548–26560, 2021.
  27. Artificial neural networks for solving ordinary and partial differential equations. IEEE transactions on neural networks, 9(5):987–1000, 1998.
  28. Non-homogeneous Boundary Value Problems and Applications: Vol. 2. Springer Science & Business Media, 1972.
  29. William Charles Hector McLean. Strongly Elliptic Systems and Boundary Integral Equations. Cambridge university press, 2000.
  30. Estimates on the generalization error of physics-informed neural networks for approximating PDEs. IMA Journal of Numerical Analysis, 43(1):1–43, 01 2022.
  31. Error estimates for the deep Ritz method with boundary penalty. In Mathematical and Scientific Machine Learning, pages 215–230. PMLR, 2022.
  32. Notes on exact boundary values in residual minimisation. In Mathematical and Scientific Machine Learning, pages 231–240. PMLR, 2022.
  33. Achieving high accuracy with PINNs via energy natural gradient descent. ICML, 2023.
  34. A PDE approach to fractional diffusion in general domains: a priori error analysis. Foundations of Computational Mathematics, 15:733–791, 2015.
  35. Least-squares mixed finite elements for second-order elliptic problems. SIAM Journal on Numerical Analysis, 31(5):1368–1377, 1994.
  36. Ab initio solution of the many-electron Schrödinger equation with deep neural networks. Physical Review Research, 2(3):033429, 2020.
  37. Physics-informed neural networks: A deep learning framework for solving forward and inverse problems involving nonlinear partial differential equations. Journal of Computational physics, 378:686–707, 2019.
  38. Neural conservation laws: A divergence-free perspective. Advances in Neural Information Processing Systems, 35:38075–38088, 2022.
  39. Martin Schechter. On Lpsuperscript𝐿𝑝{L}^{p}italic_L start_POSTSUPERSCRIPT italic_p end_POSTSUPERSCRIPT estimates and regularity II. Mathematica Scandinavica, 13(1):47–69, 1963.
  40. On the convergence of physics informed neural networks for linear second-order elliptic and parabolic type PDEs. arXiv preprint arXiv:2004.01806, 2020.
  41. Error estimates of residual minimization using neural networks for linear PDEs. arXiv preprint arXiv:2010.08019, 2020.
  42. Greedy training algorithms for neural networks and applications to PDEs. Journal of Computational Physics, 484:112084, 2023.
  43. Approximation rates for neural networks with general activation functions. Neural Networks, 128:313–321, 2020.
  44. High-order approximation rates for shallow neural networks with cosine and ReLUk activation functions. Applied and Computational Harmonic Analysis, 58:1–26, 2022.
  45. N Sukumar and Ankit Srivastava. Exact imposition of boundary conditions with distance functions in physics-informed deep neural networks. Computer Methods in Applied Mechanics and Engineering, 389:114333, 2022.
  46. Roger Temam. Navier-Stokes equations: Theory and Numerical Analysis, volume 343. American Mathematical Soc., 2001.
  47. Understanding and mitigating gradient flow pathologies in physics-informed neural networks. SIAM Journal on Scientific Computing, 43(5):A3055–A3081, 2021.
  48. When and why PINNs fail to train: A neural tangent kernel perspective. Journal of Computational Physics, 449:110768, 2022.
  49. Peter Weidemaier. Maximal regularity for parabolic equations with inhomogeneous boundary conditions in Sobolev spaces with mixed Lpsubscript𝐿𝑝L_{p}italic_L start_POSTSUBSCRIPT italic_p end_POSTSUBSCRIPT-norm. Electronic Research Announcements of the American Mathematical Society, 8(6):47–51, 2002.
  50. A comprehensive study of non-adaptive and residual-based adaptive sampling for physics-informed neural networks. Computer Methods in Applied Mechanics and Engineering, 403:115671, 2023.
  51. Jinchao Xu. The finite neuron method and convergence analysis. arXiv preprint arXiv:2010.01458, 2020.
  52. Bing Yu et al. The deep Ritz method: a deep learning-based numerical algorithm for solving variational problems. Communications in Mathematics and Statistics, 6(1):1–12, 2018.
Citations (5)
View on Semantic Scholar

Summary

We haven't generated a summary for this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Summarize Now