RUS  ENG
Full version
JOURNALS // Computer Research and Modeling // Archive

Computer Research and Modeling, 2020 Volume 12, Issue 2, Pages 301–317 (Mi crm786)

This article is cited in 6 papers

NUMERICAL METHODS AND THE BASIS FOR THEIR APPLICATION

Mirror descent for constrained optimization problems with large subgradient values of functional constraints

F. S. Stonyakinab, A. N. Stepanova, A. V. Gasnikovcdb, A. A. Titovb

a V. I. Vernadsky Crimean Federal University, 4 Academician Vernadsky av., Simferopol, 295007, Russia
b Moscow Institute of Physics and Technology, 9 Institutskiy per., Dolgoprudny, Moscow Region, 141701, Russia
c Institute for Information Transmission Problems, 19/1 Bolshoi Karetny per., Moscow, 127051, Russia
d Caucasus Mathematical Center, Adyghe State University, 208 Pervomayskaya st., Maykop, 385000, Russia

Abstract: The paper is devoted to the problem of minimization of the non-smooth functional $f$ with a non-positive non-smooth Lipschitz-continuous functional constraint. We consider the formulation of the problem in the case of quasi-convex functionals. We propose new strategies of step-sizes and adaptive stopping rules in Mirror Descent for the considered class of problems. It is shown that the methods are applicable to the objective functionals of various levels of smoothness. Applying a special restart technique to the considered version of Mirror Descent there was proposed an optimal method for optimization problems with strongly convex objective functionals. Estimates of the rate of convergence for the considered methods are obtained depending on the level of smoothness of the objective functional. These estimates indicate the optimality of the considered methods from the point of view of the theory of lower oracle bounds. In particular, the optimality of our approach for Hölder-continuous quasi-convex (sub)differentiable objective functionals is proved. In addition, the case of a quasi-convex objective functional and functional constraint was considered. In this paper, we consider the problem of minimizing a non-smooth functional $f$ in the presence of a Lipschitz-continuous non-positive non-smooth functional constraint $g$, and the problem statement in the cases of quasi-convex and strongly (quasi-)convex functionals is considered separately. The paper presents numerical experiments demonstrating the advantages of using the considered methods.

Keywords: non-smooth constrained optimization, quasi-convex functional, adaptive mirror descent, level of smoothness, Hölder-continuous objective functional, optimal method.

UDC: 519.85

Received: 04.11.2019
Revised: 08.12.2019
Accepted: 24.12.2019

Language: English

DOI: 10.20537/2076-7633-2020-12-2-301-317



© Steklov Math. Inst. of RAS, 2024