Papers
Topics
Authors
Recent
Gemini 2.5 Flash
Gemini 2.5 Flash
144 tokens/sec
GPT-4o
8 tokens/sec
Gemini 2.5 Pro Pro
46 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

Quantum Langevin Dynamics for Optimization (2311.15587v3)

Published 27 Nov 2023 in quant-ph, cs.DS, cs.LG, and math.OC

Abstract: We initiate the study of utilizing Quantum Langevin Dynamics (QLD) to solve optimization problems, particularly those non-convex objective functions that present substantial obstacles for traditional gradient descent algorithms. Specifically, we examine the dynamics of a system coupled with an infinite heat bath. This interaction induces both random quantum noise and a deterministic damping effect to the system, which nudge the system towards a steady state that hovers near the global minimum of objective functions. We theoretically prove the convergence of QLD in convex landscapes, demonstrating that the average energy of the system can approach zero in the low temperature limit with an exponential decay rate correlated with the evolution time. Numerically, we first show the energy dissipation capability of QLD by retracing its origins to spontaneous emission. Furthermore, we conduct detailed discussion of the impact of each parameter. Finally, based on the observations when comparing QLD with classical Fokker-Plank-Smoluchowski equation, we propose a time-dependent QLD by making temperature and $\hbar$ time-dependent parameters, which can be theoretically proven to converge better than the time-independent case and also outperforms a series of state-of-the-art quantum and classical optimization algorithms in many non-convex landscapes.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (53)
  1. J. Nocedal and S. J. Wright, Numerical optimization (Springer, 1999).
  2. Y. LeCun, Y. Bengio, and G. Hinton, Deep learning, nature 521, 436 (2015).
  3. S. Ruder, An overview of gradient descent optimization algorithms, arXiv preprint arXiv:1609.04747  (2016).
  4. W. Su, S. Boyd, and E. Candes, A differential equation for modeling Nesterov’s accelerated gradient method: theory and insights, Advances in neural information processing systems 27 (2014).
  5. A. Wibisono, A. C. Wilson, and M. I. Jordan, A variational perspective on accelerated methods in optimization, proceedings of the National Academy of Sciences 113, E7351 (2016).
  6. M. I. Jordan, Dynamical, symplectic and stochastic perspectives on gradient-based optimization, in Proceedings of the International Congress of Mathematicians: Rio de Janeiro 2018 (World Scientific, 2018) pp. 523–549.
  7. B. Shi, W. J. Su, and M. I. Jordan, On learning rates and Schrödinger operators, arXiv preprint arXiv:2004.06977  (2020).
  8. M. Welling and Y. W. Teh, Bayesian learning via stochastic gradient Langevin dynamics, in Proceedings of the 28th international conference on machine learning (ICML-11) (2011) pp. 681–688.
  9. S. Ahn, A. Korattikara, and M. Welling, Bayesian posterior sampling via stochastic gradient Fisher scoring, arXiv preprint arXiv:1206.6380  (2012).
  10. T. Chen, E. Fox, and C. Guestrin, Stochastic gradient hamiltonian monte carlo, in International conference on machine learning (PMLR, 2014) pp. 1683–1691.
  11. Y. Zhang, P. Liang, and M. Charikar, A hitting time analysis of stochastic gradient Langevin dynamics, in Conference on Learning Theory (PMLR, 2017) pp. 1980–2022.
  12. S. Chewi, Log-concave sampling, Book draft available at https://chewisinho. github. io  (2022).
  13. T. Li, S. Chakrabarti, and X. Wu, Sublinear quantum algorithms for training linear and kernel-based classifiers, in International Conference on Machine Learning (PMLR, 2019) pp. 3815–3824.
  14. J. van Apeldoorn and A. Gilyén, Quantum algorithms for zero-sum games, arXiv preprint arXiv:1904.03180  (2019a).
  15. P. A. Casares and M. A. Martin-Delgado, A quantum interior-point predictor–corrector algorithm for linear programming, Journal of physics A: Mathematical and Theoretical 53, 445305 (2020).
  16. F. G. Brandão and K. M. Svore, Quantum speed-ups for solving semidefinite programs, in 2017 IEEE 58th Annual Symposium on Foundations of Computer Science (FOCS) (IEEE, 2017) pp. 415–426.
  17. I. Kerenidis and A. Prakash, A quantum interior point method for LPs and SDPs, ACM Transactions on Quantum Computing 1, 1 (2020).
  18. C. Zhang, J. Leng, and T. Li, Quantum algorithms for escaping from saddle points, Quantum 5, 529 (2021).
  19. Y. Liu, W. J. Su, and T. Li, On quantum speedups for nonconvex optimization via quantum tunneling walks, Quantum 7, 1030 (2023).
  20. D. S. Ray, Notes on Brownian motion and related phenomena, arXiv preprint physics/9903033  (1999).
  21. C. W. Gardiner et al., Handbook of stochastic methods, Vol. 3 (springer Berlin, 1985).
  22. C. Gardiner and P. Zoller, Quantum noise: a handbook of Markovian and non-Markovian quantum stochastic methods with applications to quantum optics (Springer Science & Business Media, 2004).
  23. G. Ford, M. Kac, and P. Mazur, Statistical mechanics of assemblies of coupled oscillators, Journal of Mathematical Physics 6, 504 (1965).
  24. J. J. Sakurai and E. D. Commins, Modern quantum mechanics, revised edition (1995).
  25. A. Lampo, M. Á. G. March, and M. Lewenstein, Quantum Brownian motion revisited: extensions and applications (Springer, 2019).
  26. A. O. Caldeira and A. J. Leggett, Path integral approach to quantum Brownian motion, Physica A: Statistical mechanics and its Applications 121, 587 (1983a).
  27. S. Gao, Dissipative quantum dynamics with a Lindblad functional, Physical review letters 79, 3101 (1997).
  28. L. Diósi, On high-temperature Markovian equation for quantum Brownian motion, Europhysics Letters 22, 1 (1993).
  29. R. Karrlein and H. Grabert, Exact time evolution and master equations for the damped harmonic oscillator, Physical Review E 55, 153 (1997).
  30. B. Shi, On the hyperparameters in stochastic gradient descent with momentum, arXiv preprint arXiv:2108.03947  (2021).
  31. G. Ford and M. Kac, On the quantum Langevin equation, Journal of statistical physics 46, 803 (1987).
  32. G. Lindblad, On the generators of quantum dynamical semigroups, Communications in Mathematical Physics 48, 119 (1976).
  33. H. Goldstein, C. Poole, and J. Safko, Classical mechanics (2002).
  34. A. Kudlis, I. Iorsh, and I. Tokatly, Dissipation and spontaneous emission in quantum electrodynamical density functional theory based on optimized effective potential: A proof of concept study, Physical Review B 105, 054317 (2022).
  35. S. Gao, Lindblad approach to quantum dynamics of open systems, Physical Review B 57, 4509 (1998).
  36. A. O. Caldeira and A. J. Leggett, Quantum tunnelling in a dissipative system, Annals of physics 149, 374 (1983b).
  37. E. G. Harris, Quantum tunneling in dissipative systems, Physical Review A 48, 995 (1993).
  38. S. Chakravarty and A. J. Leggett, Dynamics of the two-state system with ohmic dissipation, Physical review letters 52, 5 (1984).
  39. M. P. Fisher and A. T. Dorsey, Dissipative quantum tunneling in a biased double-well system at finite temperatures, Physical review letters 54, 1609 (1985).
  40. P. Hanggi, Escape from a metastable state, Journal of Statistical Physics 42, 105 (1986).
  41. N. Kelkar, D. L. Gómez, and E. J. Patino, Time in dissipative tunneling: subtleties and applications, Annals of Physics 382, 11 (2017).
  42. D. Dolgitzer, D. Zeng, and Y. Chen, Dynamical quantum phase transitions in the spin-boson model, Optics Express 29, 23988 (2021).
  43. A. Layeb, New hard benchmark functions for global optimization, arXiv preprint arXiv:2202.04606  (2022).
  44. M. Jamil and X.-S. Yang, A literature survey of benchmark functions for global optimisation problems, International Journal of Mathematical Modelling and Numerical Optimisation 4, 150 (2013).
  45. F. J. Richards, A flexible growth function for empirical use, Journal of experimental Botany 10, 290 (1959).
  46. D. J. Griffiths and D. F. Schroeter, Introduction to quantum mechanics (Cambridge university press, 2018).
  47. J. Cohen, A. Khan, and C. Alexander, Portfolio optimization of 60 stocks using classical and quantum algorithms, arXiv preprint arXiv:2008.08669  (2020).
  48. P. Date and T. Potok, Adiabatic quantum linear regression, Scientific reports 11, 21905 (2021).
  49. Y. E. Nesterov, A method of solving a convex programming problem with convergence rate O⁢(1/k2)𝑂1superscript𝑘2O(1/k^{2})italic_O ( 1 / italic_k start_POSTSUPERSCRIPT 2 end_POSTSUPERSCRIPT ), in Doklady Akademii Nauk, Vol. 269 (Russian Academy of Sciences, 1983) pp. 543–547.
  50. W. H. Press, Numerical recipes 3rd edition: The art of scientific computing (Cambridge university press, 2007).
  51. S. Gao, D. Busch, and W. Ho, Femtosecond dynamics of electron-vibrational heating and desorption, Surface science 344, L1252 (1995).
  52. X. Li and C. Wang, Simulating Markovian open quantum systems using higher-order series expansion, arXiv preprint arXiv:2212.02051  (2022).
  53. J. Leng, Y. Zheng, and X. Wu, A quantum-classical performance separation in nonconvex optimization, arXiv preprint arXiv:2311.00811  (2023b).
Citations (3)
View on Semantic Scholar

Summary

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

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