ETRI Journal

  • 제46권3호
  • /
  • Pages.367-378
  • /
  • 2024
  • /
  • 1225-6463(pISSN)
  • /
  • 2233-7326(eISSN)
한국전자통신연구원 (Electronics and Telecommunications Research Institute)
권호 표지
DOI QR코드

DOI QR Code

Numerical analysis of quantization-based optimization

  • Jinwuk Seok (Artificial Intelligence Research Laboratory, Electronics and Telecommunications Research Institute) ;
  • Chang Sik Cho (Artificial Intelligence Research Laboratory, Electronics and Telecommunications Research Institute)
  • 투고 : 2023.02.28
  • 심사 : 2023.08.09
  • 발행 : 2024.06.20
PDF 다운로드
⟨ 이전 논문 다음 논문 ⟩

초록

We propose a number-theory-based quantized mathematical optimization scheme for various NP-hard and similar problems. Conventional global optimization schemes, such as simulated and quantum annealing, assume stochastic properties that require multiple attempts. Although our quantization-based optimization proposal also depends on stochastic features (i.e., the white-noise hypothesis), it provides a more reliable optimization performance. Our numerical analysis equates quantization-based optimization to quantum annealing, and its quantization property effectively provides global optimization by decreasing the measure of the level sets associated with the objective function. Consequently, the proposed combinatorial optimization method allows the removal of the acceptance probability used in conventional heuristic algorithms to provide a more effective optimization. Numerical experiments show that the proposed algorithm determines the global optimum in less operational time than conventional schemes.

키워드

과제정보

This work was supported by the Institute for Information and Communications Technology Promotion (IITP) grant funded by the Korean Government (MSIP) (2021-0-00766, Development of Integrated Development Framework that supports Automatic Neural Network Generation and Deployment optimized for Runtime Environment).

참고문헌

  1. G. Gutin, A. Yeo, and A. Zverovich, Traveling salesman should not be greedy: Domination analysis of greedy-type heuristics for the tsp, Discr. Appl. Math. 117 (2002), no. 1, 81-86.
  2. D. S. Johnson and L. A. McGeoch, The traveling salesman problem: A case study in local optimization, Local search in combinatorial optimization, E. H. L. Aarts and J. K. Lenstra, (eds.), John Wiley and Sons, Chichester, United Kingdom, 1997, pp. 215-310.
  3. R. Nemani, N. Cherukuri, G. R. K. Rao, P. V. V. S. Srinivas, J. J. Pujari, and C. Prasad, Algorithms and optimization techniques for solving TSP, (Fifth International Conference on I-Smac (Iot in Social, Mobile, Analytics and Cloud) (I-Smac), Palladam, India), 2021, pp. 809-814.
  4. P. Benioff, Computer as a physical system: a microscopic quantum mechanical hamiltonian model of computers represented by turing machines, J. Stat. Phys. 22 (1980), no. 5, 563-591.
  5. J. H. Holland, Adaptation in natural and artificial systems: An introductory analysis with applications to biology, control, and artificial intelligence, 1st ed., MIT Press, 1975.
  6. N. Metropolis, A. W. Rosenbluth, M. N. Rosenbluth, A. H. Teller, and E. Teller, Equation of state calculations by fast computing machines, J. Chem. Phys. 21 (1953), 1087-1092.
  7. G. Nicolis and I. Prigogine, Self-organization in nonequilibrium systems. From dissipative structures to order through fluctuations, 1st ed., Wiley and Sons, 1977.
  8. M. Pincus, A monte carlo method for the approximate solution of certain types of constrained optimization problems, Operat. Res. 18 (1970), no. 6, 1225-1228.
  9. S. Kirkpatrick, C. D. Gelatt, and M. P. Vecchi, Optimization by simulated annealing, Science 220 (1983), no. 4598, 671-680.
  10. P. J. M. Laarhoven and E. H. L. Aarts, Simulated annealing: Theory and applications, Kluwer Academic Publishers, 1987.
  11. T.-S. Chiang, C.-R. Hwang, and S. J. Sheu, Diffusion for global optimization in Rn, SIAM J. Control Optim. 25 (1987), no. 3, 737-753.
  12. S. Geman and C.-R. Hwang, Diffusions for global optimization, SIAM J. Control Optim. 24 (1986), no. 5, 1031-1043.
  13. C.-R. Hwang, Laplace's method revisited: Weak convergence of probability measures, Ann. Probab. 8 (1980), no. 6, 1177-1182.
  14. M. Locatelli, Convergence properties of simulated annealing for continuous global optimization, J. Appl. Probab. 33 (1996), no. 4, 1127-1140.
  15. D. E. Goldberg, Genetic algorithms in search, optimization and machine learning, 1st ed., Addison-Wesley Longman Publishing Co., Inc., USA, 1989.
  16. A. Hussain, Y. S. Muhammad, M. N. Sajid, I. Hussain, A. M. Shoukry, and S. Gani, Genetic algorithm for traveling salesman problem with modified cycle crossover operator, Comput. Intell. Neurosci. 2017 (2017), 1-17.
  17. M. Jiang, Y. P. Luo, and S. Y. Yang, Stochastic convergence analysis and parameter selection of the standard particle swarm optimization algorithm, Inform. Process. Lett. 102 (2007), no. 1, 8-16.
  18. J. Kennedy and R. Eberhart, Particle swarm optimization, (Proceedings of ICNN'95-International Conference on Neural Networks, Vol. 4), 1995, pp. 1942-1948.
  19. M. R. Bonyadi and Z. Michalewicz, Particle swarm optimization for single objective continuous space problems: A review, Evol. Comput. 25 (2017), no. 1, 1-54.
  20. M. Clerc, From theory to practice in particle swarm optimization, Handbook of swarm intelligence: Concepts, principles and applications, Springer Berlin Heidelberg, Berlin, Heidelberg, 2011, pp. 3-36.
  21. D. W. Gao, Solving fuzzy job-shop scheduling problem using DE algorithm improved by a selection mechanism, IEEE Trans. Fuzzy Syst. 28 (2020), no. 12, 3265-3275.
  22. G.-G. Wang, D. Gao, and W. Pedrycz, Solving multiobjective fuzzy job-shop scheduling problem by a hybrid adaptive differential evolution algorithm, IEEE Trans. Industr. Inform. 18 (2022), no. 12, 8519-8528.
  23. C. Tsallis and D. A. Stariolo, Generalized simulated annealing, Phys. A: Stat. Mech. Appl. 233 (1996), no. 1, 395-406. DOI 10. 1016/S0378-4371(96)00271-3 https://doi.org/10.1016/S0378-4371(96)00271-3
  24. B. Gidas, Nonstationary Markov chains and convergence of the annealing algorithm, J. Statistical Phys. 39 (1985), no. 4, 73-131.
  25. T. Guilmeau, E. Chouzenoux, and V. Elvira, Simulated annealing: A review and a new scheme, (IEEE Statistical Signal Processing Workshop (SSP), Rio de Janeiro, Brazil), 2021, pp. 101-105.
  26. B. Hajek, Cooling schedules for optimal annealing, Math. Operat. Res. 13 (1988), no. 2, 311-329.
  27. S. Rubenthaler, T. Ryden, and M. Wiktorsson, Fast simulated annealing in Rd with an application to maximum likelihood estimation in state-space models, Stochast. Process. Appl. 119 (2009), no. 6, 1912-1931.
  28. E. Zhou and X. Chen, Sequential monte carlo simulated annealing, J. Global Optim. 55 (2013), no. 1, 101-124.
  29. J. Seok and J.-S. Kim, Nonlinear optimization algorithm using monotonically increasing quantization resolution, ETRI J. 45 (2023), no. 1, 119-130. DOI 10.4218/etrij.2021-0320
  30. R. Biswas, Z. Jiang, K. Kechezhi, S. Knysh, S. Mandra, B. O'Gorman, A. Perdomo-Ortiz, A. Petukhov, J. Realpe-Gomez, E. Rieffel, D. Venturelli, F. Vasko, and Z. Wang, A nasa perspective on quantum computing: Opportunities and challenges, Parallel Comput. 64 (2017), 81-98.
  31. E. Crosson and A. W. Harrow, Simulated quantum annealing can be exponentially faster than classical simulated annealing, (IEEE 57th Annual Symposium on Foundations of Computer Science (FOCS), New Brunswick, NJ, USA), 2016, pp. 714-723.
  32. R. C. Foster, B. Weaver, and J. R. Gattiker, Applications of quantum annealing in statistics, 2019. arXiv: 1904.06819.
  33. D. de Falco and D. Tamascelli, An introduction to quantum annealing, RAIRO-Theoret. Inform. Appl. 45 (2011), no. 1, 99-116.
  34. G. E. Santoro and E. Tosatti, Optimization using quantum mechanics: Quantum annealing through adiabatic evolution, J. Phys. A: Math. Gen. 39 (2006), no. 36, R393-R431.
  35. L. Stella, G. E. Santoro, and E. Tosatti, Optimization by quantum annealing: Lessons from simple cases, Phys. Rev. B 72 (2005), no. 1, 14303.
  36. B. Heim, T. F. Ronnow, S. V. Isakov, and M. Troyer, Quantum versus classical annealing of Ising spin glasses, Science 348 (2015), no. 6231, 215-217.
  37. J. Seok and C. Cho, Quantization-based optimization: Alternative stochastic approximation of global optimization, (OPT2022: 14th Annual Workshop on Optimization for Machine Learning in NeurIPS 2022), 2022, pp. 1-25.
  38. F. C. Klebaner, Introduction to stochastic calculus with applications, 3rd ed., Imperial College Press, 2012.
  39. B. Oksendal, Stochastic differential equations: An introduction with applications, 6th ed., Springer-Verlag, 2003.