Computer Research and Modeling
RUS  ENG    JOURNALS   PEOPLE   ORGANISATIONS   CONFERENCES   SEMINARS   VIDEO LIBRARY   PACKAGE AMSBIB  
General information
Latest issue
Archive

Search papers
Search references

RSS
Latest issue
Current issues
Archive issues
What is RSS



Computer Research and Modeling:
Year:
Volume:
Issue:
Page:
Find






Personal entry:
Login:
Password:
Save password
Enter
Forgotten password?
Register


Computer Research and Modeling, 2022, Volume 14, Issue 2, Pages 239–255
DOI: https://doi.org/10.20537/2076-7633-2022-14-2-239-255
(Mi crm966)
 

This article is cited in 1 scientific paper (total in 1 paper)

MATHEMATICAL MODELING AND NUMERICAL SIMULATION

Linearly convergent gradient-free methods for minimization of parabolic approximation

A. I. Bazarovaa, A. N. Beznosikova, A. V. Gasnikovabc

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 Caucasus Mathematical Center, Adyghe State University, Maikop
Full-text PDF (510 kB) Citations (1)
References:
Abstract: Finding the global minimum of a nonconvex function is one of the key and most difficult problems of the modern optimization. In this paper we consider special classes of nonconvex problems which have a clear and distinct global minimum.
In the first part of the paper we consider two classes of «good» nonconvex functions, which can be bounded below and above by a parabolic function. This class of problems has not been widely studied in the literature, although it is rather interesting from an applied point of view. Moreover, for such problems first-order and higher-order methods may be completely ineffective in finding a global minimum. This is due to the fact that the function may oscillate heavily or may bevery noisy. Therefore, our new methods use only zero-order information and are based on grid search. The size and fineness of this grid, and hence the guarantee of convergence speed and oracle complexity, depend on the «goodness» of the problem. In particular, we show that if the function is bounded by fairly close parabolic functions, then the complexity is independent of the dimension of the problem. We show that our new methods converge with a linear convergence rate $\log(1/\epsilon)$ to a global minimum on the cube.
In the second part of the paper, we consider the nonconvex optimization problem from a different angle. We assume that the target minimizing function is the sum of the convex quadratic problem and a nonconvex «noise» function proportional to the distance to the global solution. Considering functions with such noise assumptions for zero-order methods is new in the literature. For such a problem, we use the classical gradient-free approach with gradient approximation through finite differences. We show how the convergence analysis for our problems can be reduced to the standard analysis for convex optimization problems. In particular, we achieve a linear convergence rate for such problems as well.
Experimental results confirm the efficiency and practical applicability of all the obtained methods.
Keywords: zeroth-order optimization, nonconvex problem, linear rate.
Funding agency Grant number
Russian Science Foundation 21-71-30005
This research was funded by the Russian Science Foundation (project 21-71-30005).
Received: 09.02.2022
Accepted: 13.02.2022
Document Type: Article
UDC: 519.8
Language: English
Citation: A. I. Bazarova, A. N. Beznosikov, A. V. Gasnikov, “Linearly convergent gradient-free methods for minimization of parabolic approximation”, Computer Research and Modeling, 14:2 (2022), 239–255
Citation in format AMSBIB
\Bibitem{BazBezGas22}
\by A.~I.~Bazarova, A.~N.~Beznosikov, A.~V.~Gasnikov
\paper Linearly convergent gradient-free methods for minimization of parabolic approximation
\jour Computer Research and Modeling
\yr 2022
\vol 14
\issue 2
\pages 239--255
\mathnet{http://mi.mathnet.ru/crm966}
\crossref{https://doi.org/10.20537/2076-7633-2022-14-2-239-255}
Linking options:
  • https://www.mathnet.ru/eng/crm966
  • https://www.mathnet.ru/eng/crm/v14/i2/p239
  • This publication is cited in the following 1 articles:
    Citing articles in Google Scholar: Russian citations, English citations
    Related articles in Google Scholar: Russian articles, English articles
    Computer Research and Modeling
    Statistics & downloads:
    Abstract page:97
    Full-text PDF :27
    References:16
     
      Contact us:
     Terms of Use  Registration to the website  Logotypes © Steklov Mathematical Institute RAS, 2024