|
This article is cited in 13 scientific papers (total in 13 papers)
Optimal control
Accelerated meta-algorithm for convex optimization problems
A. V. Gasnikovab, D. M. Dvinskikhabc, P. E. Dvurechenskiibc, D. Kamzolova, V. V. Matyukhina, D. A. Pasechnyuka, N. K. Tupitsaa, A. V. Chernova a Moscow Institute of Physics and Technology (National Research University), Dolgoprudny, Moscow Region
b Institute for Information Transmission Problems of the Russian Academy of Sciences (Kharkevich Institute), Moscow
c Weierstrass institute for Applied Analysis and Stochastics
Abstract:
An envelope called an accelerated meta-algorithm is proposed. Based on the envelope, accelerated methods for solving convex unconstrained minimization problems in various formulations can be obtained from nonaccelerated versions in a unified manner. Quasi-optimal algorithms for minimizing smooth functions with Lipschitz continuous derivatives of arbitrary order and for solving smooth minimax problems are given as applications. The proposed envelope is more general than existing ones. Moreover, better convergence estimates can be obtained in the case of this envelope and better efficiency can be achieved in practice for a number of problem formulations.
Key words:
convex optimization, accelerated proximal method, tensor methods, inexact oracle, sliding, catalyst.
Received: 18.04.2020 Revised: 16.06.2020 Accepted: 18.09.2020
Citation:
A. V. Gasnikov, D. M. Dvinskikh, P. E. Dvurechenskii, D. Kamzolov, V. V. Matyukhin, D. A. Pasechnyuk, N. K. Tupitsa, A. V. Chernov, “Accelerated meta-algorithm for convex optimization problems”, Zh. Vychisl. Mat. Mat. Fiz., 61:1 (2021), 20–31; Comput. Math. Math. Phys., 61:1 (2021), 17–28
Linking options:
https://www.mathnet.ru/eng/zvmmf11181 https://www.mathnet.ru/eng/zvmmf/v61/i1/p20
|
Statistics & downloads: |
Abstract page: | 154 | References: | 34 |
|