Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
129 tokens/sec
GPT-4o
28 tokens/sec
Gemini 2.5 Pro Pro
42 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

Investigating the Ability of PINNs To Solve Burgers' PDE Near Finite-Time BlowUp (2310.05169v2)

Published 8 Oct 2023 in cs.LG, cs.NA, math.AP, and math.NA

Abstract: Physics Informed Neural Networks (PINNs) have been achieving ever newer feats of solving complicated PDEs numerically while offering an attractive trade-off between accuracy and speed of inference. A particularly challenging aspect of PDEs is that there exist simple PDEs which can evolve into singular solutions in finite time starting from smooth initial conditions. In recent times some striking experiments have suggested that PINNs might be good at even detecting such finite-time blow-ups. In this work, we embark on a program to investigate this stability of PINNs from a rigorous theoretical viewpoint. Firstly, we derive generalization bounds for PINNs for Burgers' PDE, in arbitrary dimensions, under conditions that allow for a finite-time blow-up. Then we demonstrate via experiments that our bounds are significantly correlated to the $\ell_2$-distance of the neurally found surrogate from the true blow-up solution, when computed on sequences of PDEs that are getting increasingly close to a blow-up.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (50)
  1. Jameson A, Martinelli L and Vassberg J 2002 Using computational fluid dynamics for aerodynamics–a critical assessment Proceedings of ICAS pp 2002–1
  2. Lagaris I, Likas A and Fotiadis D 1998 IEEE Transactions on Neural Networks 9 987–1000 URL https://doi.org/10.1109%2F72.712178
  3. Broomhead D and Lowe D 1988 ROYAL SIGNALS AND RADAR ESTABLISHMENT MALVERN (UNITED KINGDOM) RSRE-MEMO-4148
  4. Karniadakis, GE and Kevrekidis 2021 Physics-informed machine learning
  5. E W, Han J and Jentzen A 2021 Nonlinearity 35 278–310 URL https://doi.org/10.1088%2F1361-6544%2Fac337f
  6. Raissi M, Perdikaris P and Karniadakis G E 2019 Journal of Computational physics 378 686–707
  7. Yu B et al. 2018 Communications in Mathematics and Statistics 6 1–12
  8. Sirignano J and Spiliopoulos K 2018 Journal of computational physics 375 1339–1364
  9. Kaiser E, Kutz J N and Brunton S L 2021 Data-driven discovery of koopman eigenfunctions for control (Preprint 1707.01146)
  10. Erichson N B, Muehlebach M and Mahoney M W 2019 Physics-informed autoencoders for lyapunov-stable fluid flow prediction (Preprint 1905.10866)
  11. Wandel N, Weinmann M and Klein R 2021 Physics of Fluids 33 047117 URL https://doi.org/10.1063%2F5.0047428
  12. Salvi C, Lemercier M and Gerasimovics A 2022 Neural stochastic pdes: Resolution-invariant learning of continuous spatiotemporal dynamics (Preprint 2110.10249)
  13. Wang S, Wang H and Perdikaris P 2021 Science Advances 7 eabi8605 (Preprint https://www.science.org/doi/pdf/10.1126/sciadv.abi8605) URL https://www.science.org/doi/abs/10.1126/sciadv.abi8605
  14. Arthurs C J and King A P 2021 Journal of Computational Physics 438 110364 ISSN 0021-9991 URL https://www.sciencedirect.com/science/article/pii/S002199912100259X
  15. Mishra S and Molinaro R 2022 IMA Journal of Numerical Analysis 43 1–43 ISSN 0272-4979 (Preprint https://academic.oup.com/imajna/article-pdf/43/1/1/49059512/drab093.pdf) URL https://doi.org/10.1093/imanum/drab093
  16. De Ryck T, Jagtap A D and Mishra S 2022 arXiv preprint arXiv:2203.09346
  17. Wang S, Teng Y and Perdikaris P 2021 SIAM Journal on Scientific Computing 43 A3055–A3081 (Preprint https://doi.org/10.1137/20M1318043) URL https://doi.org/10.1137/20M1318043
  18. Wang S, Sankaran S and Perdikaris P 2022 arXiv preprint arXiv:2203.07404
  19. Wintner A 1945 American Journal of Mathematics 67 277–284 ISSN 00029327, 10806377 URL http://www.jstor.org/stable/2371729
  20. Cooke K L 1955 Rendiconti del Circolo Matematico di Palermo 4 301–308 ISSN 1973-4409 URL https://doi.org/10.1007/BF02854201
  21. Pazy A 1983 Some Nonlinear Evolution Equations (New York, NY: Springer New York) pp 183–205 ISBN 978-1-4612-5561-1 URL https://doi.org/10.1007/978-1-4612-5561-1_6
  22. Lin Y, Sontag E D and Wang Y 1996 SIAM Journal on Control and Optimization 34 124–160 (Preprint https://doi.org/10.1137/S0363012993259981) URL https://doi.org/10.1137/S0363012993259981
  23. Stuart A M and Floater M S 1990 European Journal of Applied Mathematics 1 47–71
  24. Fujita H 1966 On the blowing up of solutions of the cauchy problem for u1+δ⁢u+u1+αsubscript𝑢1𝛿𝑢superscript𝑢1𝛼u_{1}+\delta u+u^{1+\alpha}italic_u start_POSTSUBSCRIPT 1 end_POSTSUBSCRIPT + italic_δ italic_u + italic_u start_POSTSUPERSCRIPT 1 + italic_α end_POSTSUPERSCRIPT URL https://api.semanticscholar.org/CorpusID:118871869
  25. Fujita H 1969 Bulletin of the American Mathematical Society 75 132 – 135
  26. Herrero M A and Velazquez J J L 1997 Annali della Scuola Normale Superiore di Pisa - Classe di Scienze Ser. 4, 24 633–683 URL http://www.numdam.org/item/ASNSP_1997_4_24_4_633_0/
  27. He S and Tadmor E 2019 Archive for Rational Mechanics and Analysis 232 951–986
  28. Chen J, Hou T Y and Huang D 2022 Annals of PDE 8 24
  29. Tanaka Y 2023 arXiv preprint arXiv:2304.13421
  30. Bebernes J and Kassoy D 1981 SIAM Journal on Applied Mathematics 40 476–484
  31. Lacey A 1983 SIAM Journal on Applied Mathematics 43 1350–1366
  32. Dold J 1991 Proceedings of the Royal Society of London. Series A: Mathematical and Physical Sciences 433 521–545
  33. Herrero M A and Velázquez J 1993 Israel Journal of Mathematics 81 321–341
  34. Lacey A A 1995 European Journal of Applied Mathematics 6 201–224
  35. Tao T 2016 Annals of PDE 2 1–79
  36. Tao T 2016 Journal of the American Mathematical Society 29 601–674
  37. Wang S, Yu X and Perdikaris P 2022 Journal of Computational Physics 449 110768 ISSN 0021-9991 URL https://www.sciencedirect.com/science/article/pii/S002199912100663X
  38. Johnson C and Szepessy A 1987 Mathematics of Computation - Math. Comput. 49 427–427
  39. Tadmor E 1991 SIAM Journal on Numerical Analysis 28 891–906 (Preprint https://doi.org/10.1137/0728048) URL https://doi.org/10.1137/0728048
  40. Nessyahu H and Tadmor E 1992 SIAM Journal on Numerical Analysis 29 1505–1519 (Preprint https://doi.org/10.1137/0729087) URL https://doi.org/10.1137/0729087
  41. Dziugaite G K and Roy D M 2017 arXiv preprint arXiv:1703.11008
  42. Neyshabur B, Bhojanapalli S and Srebro N 2017 arXiv preprint arXiv:1707.09564
  43. Mukherjee A 2020 A study of the mathematics of deep learning Ph.D. thesis The Johns Hopkins University
  44. Muthukumar R and Sulam J 2023 Sparsity-aware generalization theory for deep neural networks Proceedings of Thirty Sixth Conference on Learning Theory (Proceedings of Machine Learning Research vol 195) ed Neu G and Rosasco L (PMLR) pp 5311–5342 URL https://proceedings.mlr.press/v195/muthukumar23a.html
  45. Biazar J and Aminikhah H 2009 Mathematical and Computer Modelling 49 1394–1400
  46. Luo G and Hou T Y 2014 Proceedings of the National Academy of Sciences 111 12968–12973 (Preprint https://www.pnas.org/doi/pdf/10.1073/pnas.1405238111) URL https://www.pnas.org/doi/abs/10.1073/pnas.1405238111
  47. Elgindi T M 2021 Annals of Mathematics 194 647 – 727 URL https://doi.org/10.4007/annals.2021.194.3.2
  48. Chen J and Hou T Y 2021 Communications in Mathematical Physics 383 1559–1667
  49. Luo G and Hou T Y 2014 Multiscale Modeling & Simulation 12 1722–1776 (Preprint https://doi.org/10.1137/140966411) URL https://doi.org/10.1137/140966411
  50. DeVore R A and Scott L R 1984 SIAM Journal on Numerical Analysis 21 400–412
Citations (2)
View on Semantic Scholar

Summary

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

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