Adaptive step adjustment for a stochastic optimization algorithm

F. Mirzoakhmedov; S. P. Uryas'ev

doi:10.1016/S0041-5553(83)80071-8

Adaptive step adjustment for a stochastic optimization algorithm

F. Mirzoakhmedov
, S. P. Uryas'ev

Applied Mathematics and Statistics

Research output: Contribution to journal › Article › peer-review

Abstract

The quasigradient algorithm of stochastic optimization is considered. The conditions to be imposed on the step multiplier, for Cesaro convergence of the algorithms with probability 1, are studied. Adaptive step adjustment is proposed, and the convergence of the corresponding algorithm is proved. A numerical algorithm containing heuristic elements is described. The results of numerical experiments are quoted.

Original language	English
Pages (from-to)	20-27
Number of pages	8
Journal	USSR Computational Mathematics and Mathematical Physics
Volume	23
Issue number	6
DOIs	https://doi.org/10.1016/S0041-5553(83)80071-8
State	Published - 1983

Access to Document

10.1016/S0041-5553(83)80071-8

Cite this

@article{19d3a0f4a457434d93a108757c7a923f,

title = "Adaptive step adjustment for a stochastic optimization algorithm",

abstract = "The quasigradient algorithm of stochastic optimization is considered. The conditions to be imposed on the step multiplier, for Cesaro convergence of the algorithms with probability 1, are studied. Adaptive step adjustment is proposed, and the convergence of the corresponding algorithm is proved. A numerical algorithm containing heuristic elements is described. The results of numerical experiments are quoted.",

author = "F. Mirzoakhmedov and Uryas'ev, \{S. P.\}",

year = "1983",

doi = "10.1016/S0041-5553(83)80071-8",

language = "English",

volume = "23",

pages = "20--27",

journal = "USSR Computational Mathematics and Mathematical Physics",

issn = "0041-5553",

number = "6",

}

TY - JOUR

T1 - Adaptive step adjustment for a stochastic optimization algorithm

AU - Mirzoakhmedov, F.

AU - Uryas'ev, S. P.

PY - 1983

Y1 - 1983

N2 - The quasigradient algorithm of stochastic optimization is considered. The conditions to be imposed on the step multiplier, for Cesaro convergence of the algorithms with probability 1, are studied. Adaptive step adjustment is proposed, and the convergence of the corresponding algorithm is proved. A numerical algorithm containing heuristic elements is described. The results of numerical experiments are quoted.

AB - The quasigradient algorithm of stochastic optimization is considered. The conditions to be imposed on the step multiplier, for Cesaro convergence of the algorithms with probability 1, are studied. Adaptive step adjustment is proposed, and the convergence of the corresponding algorithm is proved. A numerical algorithm containing heuristic elements is described. The results of numerical experiments are quoted.

UR - https://www.scopus.com/pages/publications/58149367968

U2 - 10.1016/S0041-5553(83)80071-8

DO - 10.1016/S0041-5553(83)80071-8

M3 - Article

AN - SCOPUS:58149367968

SN - 0041-5553

VL - 23

SP - 20

EP - 27

JO - USSR Computational Mathematics and Mathematical Physics

JF - USSR Computational Mathematics and Mathematical Physics

IS - 6

ER -

Adaptive step adjustment for a stochastic optimization algorithm

Abstract

Access to Document

Other files and links

Fingerprint

Cite this