Email this article Accelerated Gradient-Free Optimization Methods with a Non-Euclidean Proximal Operator
- Authors: Vorontsova E.A.1,2, Gasnikov A.V.3,4,5, Gorbunov E.A.3, Dvurechenskii P.E.6
-
Affiliations:
- Far Eastern Federal University
- Université Grenoble Alps
- Moscow Institute of Physics and Technology
- National Research University Higher School of Economics
- Caucasus Mathematical Center
- Weierstrass Institute for Applied Analysis and Stochastics
- Issue: Vol 80, No 8 (2019)
- Pages: 1487-1501
- Section: Optimization, System Analysis, and Operations Research
- URL: https://journals.rcsi.science/0005-1179/article/view/151140
- DOI: https://doi.org/10.1134/S0005117919080095
- ID: 151140
Cite item
Open Access
Access granted
Subscription Access
Abstract
We propose an accelerated gradient-free method with a non-Euclidean proximal operator associated with the p-norm (1 ⩽ p ⩽ 2). We obtain estimates for the rate of convergence of the method under low noise arising in the calculation of the function value. We present the results of computational experiments.
About the authors
E. A. Vorontsova
Far Eastern Federal University; Université Grenoble Alps
Author for correspondence.
Email: vorontsovaea@gmail.com
Russian Federation, Vladivostok; Grenoble
A. V. Gasnikov
Moscow Institute of Physics and Technology; National Research University Higher School of Economics; Caucasus Mathematical Center
Author for correspondence.
Email: gasnikov@yandex.ru
Russian Federation, Moscow; Moscow; Maikop, Republic of Adygea
E. A. Gorbunov
Moscow Institute of Physics and Technology
Author for correspondence.
Email: ed-gorbunov@yandex.ru
Russian Federation, Moscow
P. E. Dvurechenskii
Weierstrass Institute for Applied Analysis and Stochastics
Author for correspondence.
Email: pavel.dvurechensky@gmail.com
Germany, Berlin
Supplementary files
