J. Phys. Soc. Jpn. 90, 093001 (2021) [5 Pages]
LETTERS

Neural Network Approach to Construction of Classical Integrable Systems

+ Affiliations
1Department of Physics, The University of Tokyo, Bunkyo, Tokyo 113-0033, Japan2Institute for Solid State Physics, The University of Tokyo, Kashiwa, Chiba 277-8581, Japan

Integrable systems have provided various insights into physical phenomena and mathematics. The way of constructing many-body integrable systems is limited to few ansatzes for the Lax pair, except for highly inventive findings of conserved quantities. Machine learning techniques have recently been applied to broad physics fields and proven powerful for building non-trivial transformations and potential functions. We here propose a machine learning approach to a systematic construction of classical integrable systems. Given the Hamiltonian or samples in latent space, our neural network simultaneously learns the corresponding natural Hamiltonian in real space and the canonical transformation between the latent space and the real space variables. We also propose a loss function for building integrable systems and demonstrate successful unsupervised learning for the Toda lattice. Our approach enables exploring new integrable systems without any prior knowledge about the canonical transformation or any ansatz for the Lax pair.

©2021 The Physical Society of Japan

References

  • 1 M. Toda, J. Phys. Soc. Jpn. 22, 431 (1967). 10.1143/JPSJ.22.431 LinkGoogle Scholar
  • 2 M. Toda, J. Phys. Soc. Jpn. 23, 501 (1967). 10.1143/JPSJ.23.501 LinkGoogle Scholar
  • 3 F. Calogero, Lett. Nuovo Cimento (1971–1985) 13, 411 (1975). 10.1007/BF02790495 CrossrefGoogle Scholar
  • 4 F. Calogero, Lett. Nuovo Cimento (1971–1985) 16, 77 (1976). 10.1007/BF02824328 CrossrefGoogle Scholar
  • 5 J. Moser, Adv. Math. 16, 197 (1975). 10.1016/0001-8708(75)90151-6 CrossrefGoogle Scholar
  • 6 G. Carleo, I. Cirac, K. Cranmer, L. Daudet, M. Schuld, N. Tishby, L. Vogt-Maranto, and L. Zdeborová, Rev. Mod. Phys. 91, 045002 (2019). 10.1103/RevModPhys.91.045002 CrossrefGoogle Scholar
  • 7 F. Noé, S. Olsson, J. Köhler, and H. Wu, Science 365, eaaw1147 (2019). 10.1126/science.aaw1147 CrossrefGoogle Scholar
  • 8 S.-H. Li, C.-X. Dong, L. Zhang, and L. Wang, Phys. Rev. X 10, 021020 (2020). 10.1103/PhysRevX.10.021020 CrossrefGoogle Scholar
  • 9 R. Bondesan and A. Lamacraft, arXiv:1906.04645. Google Scholar
  • 10 S. Greydanus, M. Dzamba, and J. Yosinski, in Advances in Neural Information Processing Systems, ed. H. Wallach, H. Larochelle, A. Beygelzimer, F. d’Alché-Buc, E. Fox, and R. Garnett (2019) p. 15379, Vol. 32. Google Scholar
  • 11 H. Suwa, J. S. Smith, N. Lubbers, C. D. Batista, G.-W. Chern, and K. Barros, Phys. Rev. B 99, 161107 (2019). 10.1103/PhysRevB.99.161107 CrossrefGoogle Scholar
  • 12 V. Arnol’d, Mathematical Methods of Classical Mechanics (Springer, New York, 1989) 2nd ed., Vol. 60 of Graduate Texts in Mathematics. CrossrefGoogle Scholar
  • 13 A. Perelomov, Integrable Systems of Classical Mechanics and Lie Algebras (Birkhäuser, Basel, 1990) Vol. 1. CrossrefGoogle Scholar
  • 14 G. Arutyunov, Elements of Classical and Quantum Integrable Systems (Springer International Publishing, 2019) 1st ed., 2198-7882. CrossrefGoogle Scholar
  • 15 (Supplemental Material) The details of the action-angle variables, the structure of the neural networks, and the learning performance are available online. Google Scholar
  • 16 M. Cariglia, Rev. Mod. Phys. 86, 1283 (2014). 10.1103/RevModPhys.86.1283 CrossrefGoogle Scholar
  • 17 N. T. Zung, Ann. Math. 161, 141 (2005). 10.4007/annals.2005.161.141 CrossrefGoogle Scholar
  • 18 H. Ito, Comment. Math. Helv. 64, 412 (1989). 10.1007/BF02564686 CrossrefGoogle Scholar
  • 19 T. Kappeler, Y. Kodama, and A. Némethi, Annali della Scuola Normale Superiore di Pisa - Classe di Scienze Ser. 4 26, 623 (1998). Google Scholar
  • 20 A. Henrici and T. Kappeler, in Birkhoff Normal Form for the Periodic Toda Lattice, ed. J. Baik (American Mathematical Society, Providence, RI, 2008) number 458 in Contemporary Mathematics, p. 11. CrossrefGoogle Scholar
  • 21 A. Henrici, Discrete & Continuous Dynamical Systems - A 35, 2949 (2015). 10.3934/dcds.2015.35.2949 CrossrefGoogle Scholar
  • 22 K. He, X. Zhang, S. Ren, and J. Sun, IEEE Conf. Computer Vision and Pattern Recognition (CVPR), 2016, p. 770. 10.1109/CVPR.2016.90 CrossrefGoogle Scholar
  • 23 A. Sehanobish, H. Corzo, O. Kara, and D. V. Dijk, arXiv:2006.13297. Google Scholar
  • 24 K. Iwasawa, Ann. Math. 50, 507 (1949). 10.2307/1969548 CrossrefGoogle Scholar
  • 25 R. V. L. Hartley, Proc. IRE 30, 144 (1942). 10.1109/JRPROC.1942.234333 CrossrefGoogle Scholar
  • 26 R. N. Bracewell, J. Opt. Soc. Am. 73, 1832 (1983). 10.1364/JOSA.73.001832 CrossrefGoogle Scholar
  • 27 L. Dinh, J. Sohl-Dickstein, and S. Bengio, 5th Int. Conf. Learning Representations, ICLR 2017, Conference Track Proceedings, 2017. Google Scholar
  • 28 T. Teshima, I. Ishikawa, K. Tojo, K. Oono, M. Ikeda, and M. Sugiyama, in Advances in Neural Information Processing Systems, ed. H. Larochelle, M. Ranzato, R. Hadsell, M. F. Balcan, and H. Lin, Vol. 33, p. 3362. Google Scholar
  • 29 D. P. Kingma and J. Ba, arXiv:1412.6980. Google Scholar
  • 30 J. Liu, L. Sun, Q. Li, J. Ming, Y. Liu, and H. Xiong, Proc. 23rd ACM SIGKDD Int. Conf. Knowledge Discovery and Data Mining, KDD’17, 2017, p. 957. 10.1145/3097983.3098180 CrossrefGoogle Scholar
  • 31 Y. Zhou and Y. Huang, IEEE Int. Conf. Big Data (Big Data), 2019, p. 1577. 10.1109/BigData47090.2019.9006286 CrossrefGoogle Scholar
  • 32 S. Van Canneyt, P. Leroux, B. Dhoedt, and T. Demeester, Multimedia Tools Appl. 77, 1409 (2018). 10.1007/s11042-017-4348-z CrossrefGoogle Scholar
  • 33 J. M. Sanz-Serna, SIAM Rev. 58, 3 (2016). 10.1137/151002769 CrossrefGoogle Scholar
  • 34 R. T. Q. Chen, Y. Rubanova, J. Bettencourt, and D. Duvenaud, Proc. 32nd Int. Conf. Neural Information Processing Systems, NIPS’18, 2018, p. 6572. Google Scholar
  • 35 H. Flaschka and D. W. McLaughlin, Prog. Theor. Phys. 55, 438 (1976). 10.1143/PTP.55.438 CrossrefGoogle Scholar
  • 36 H. Flaschka, Prog. Theor. Phys. 51, 703 (1974). 10.1143/PTP.51.703 CrossrefGoogle Scholar
  • 37 S. Duane, A. Kennedy, B. J. Pendleton, and D. Roweth, Phys. Lett. B 195, 216 (1987). 10.1016/0370-2693(87)91197-X CrossrefGoogle Scholar
  • 38 S. Brooks, A. Gelman, G. Jones, and X.-L. Meng, Handbook of Markov Chain Monte Carlo (CRC Press, New York, 2011). CrossrefGoogle Scholar
  • 39 L. E. Reichl, The Transition to Chaos (Springer, New York, 1993). Google Scholar