Email this article Fast Gradient Descent for Convex Minimization Problems with an Oracle Producing a (δ, L)-Model of Function at the Requested Point
- Authors: Gasnikov A.V.1,2,3, Tyurin A.I.1
-
Affiliations:
- State University—Higher School of Economics
- Moscow Institute of Physics and Technology
- Kharkevich Institute for Information Transmission Problems
- Issue: Vol 59, No 7 (2019)
- Pages: 1085-1097
- Section: Article
- URL: https://journal-vniispk.ru/0965-5425/article/view/180681
- DOI: https://doi.org/10.1134/S0965542519070078
- ID: 180681
Cite item
Open Access
Access granted
Subscription Access
Abstract
A new concept of \((\delta ,L)\)-model of a function that is a generalization of the Devolder–Glineur–Nesterov \((\delta ,L)\)-oracle is proposed. Within this concept, the gradient descent and fast gradient descent methods are constructed and it is shown that constructs of many known methods (composite methods, level methods, conditional gradient and proximal methods) are particular cases of the methods proposed in this paper.
About the authors
A. V. Gasnikov
State University—Higher School of Economics; Moscow Institute of Physics and Technology; Kharkevich Institute for Information Transmission Problems
Email: atyurin@hse.ru
Russian Federation, Moscow, 125319; Dolgoprudnyi, Moscow oblast, 141700; Moscow, 127051
A. I. Tyurin
State University—Higher School of Economics
Author for correspondence.
Email: atyurin@hse.ru
Russian Federation, Moscow, 125319
Supplementary files
