Variable Metric Primal-Dual Method for Convex Optimization Problems with Changing Constraints

Igor Vasilyevich Konnov

ПРЕПРИНТ

Эта статья является препринтом и не была отрецензирована.
О результатах, изложенных в препринтах, не следует сообщать в СМИ как о проверенной информации.

Variable Metric Primal-Dual Method for Convex Optimization Problems with Changing Constraints

I. V. Konnov

2022-08-18

https://doi.org/10.24108/preprints-3112463

We propose a modified primal-dual method for general convex optimization problems with changing affine constraints. We establish convergence of the method that uses variable metric matrices at each iteration. This approach yields new opportunities for control of the parameters according to the constraints changes. In case of the multi-agent optimization problems the method can be adjusted to the changing communication topology and enables the agents to choose the parameters separately of each other.

Ссылка для цитирования:

Konnov I. V. 2022. Variable Metric Primal-Dual Method for Convex Optimization Problems with Changing Constraints. PREPRINTS.RU. https://doi.org/10.24108/preprints-3112463

Список литературы

1. Khan, M., Pandurangan, G., Kumar, V.: Distributed algorithms for constructing approximate minimum spanning trees in wireless sensor networks. IEEE Trans. Paral. Distrib. Syst. 20, 124-139 (2009)

2. Lobel, I., Ozdaglar, A., Feijer, D.: Distributed multi-agent optimization with state-dependent communication. Math. Program. 129, 255-284 (2011)

3. Peng, Z., Yan, M., Yin, W.: Parallel and distributed sparse optimization. The 47th Asilomar Conference on Signals, Systems and Computers, pp.646-659. Pacific Grove, IEEE (2013)

4. Konnov, I.V.: Primal-dual method for optimization problems with changing constraints. In: P. Pardalos et al. (eds.) Mathematical Optimization Theory and Operations Research (MOTOR 2022), Lecture Notes in Computer Science \textbf{13367}, pp. 46-61. Springer, Cham (2022)

5. Antipin, A.S.: On non-gradient methods for optimization of saddle functions. In: Karmanov, V.G. (ed.) Problems of Cybernetics. Methods and Algorithms for the Analysis of Large Systems, pp.4-13. Nauchn. Sovet po Probleme "Kibernetika", Moscow (1988) [In Russian]

6. Gol'shtein, E.G., Tret'yakov, N.V.: Modified Lagrange functions. Nauka, Moscow (1989) [Engl. transl. in John Wiley and Sons, New York (1996)].

7. Rockafellar, R.T.: Convex analysis. Princeton University Press, Princeton (1970)

8. Polyak, B.T.: Introduction to optimization. Nauka, Moscow (1983) [Engl. transl. in Optimization Software, New York (1987)]

9. Kontogiorgis, S., Meyer, R.: A variable-penalty alternating directions method for convex optimization.

10. Math. Program. 83, 29-53 (1998)

11. He, B., Wang, S., Yang, H.: A modified variable-penalty alternating directions method for monotone variational inequalities. J. Comput. Mathem. 21, 495-504 (2003)

12. Gantmacher, F.R.: The theory of matrices. Nauka, Moscow, 1966) [In Russian]

13. Hamedani, E.Y., Aybat, N.S.: A decentralized primal-dual method for constrained minimization of a strongly convex function. Preprint arXiv:1908.11835v4 on 21 Feb. 2022. - 27 pp.

14. Blondel, V.D., Hendrickx, J.M., Olshevsky, A., Tsitsiklis, J.N.: Convergence in multiagent coordination, consensus, and flocking. In: Proceedings of the 44-th IEEE Conference on Decision and Control, Seville

15. (2005) - 5 pp.

16. Nedic, A., Olshevsky, A.: Distributed optimization over time-varying directed graphs. IEEE Trans. Autom. Control. 60, 601-615 (2015)

17. Aybat, N.S., Hamedani, E.Y.: A primal-dual method for conic constrained distributed optimization problems. In: Advances in Neural Information Processing Systems, pp.5049-5057. Neural Information Processing Systems Foundation, Barcelona (2016)

ПРЕПРИНТ

ИНФОРМАЦИЯ