Dynamic stochastic optimization

Craig Wilson; Venugopal Veeravalli; Angelia Nedic

doi:10.1109/CDC.2014.7039377

Dynamic stochastic optimization

Craig Wilson, Venugopal Veeravalli, Angelia Nedic

Research output: Contribution to journal › Conference article › peer-review

3 Scopus citations

Abstract

A framework for sequentially solving stochastic optimization problems with stochastic gradient descent is introduced. Two tracking criteria are considered, one based on being accurate with respect to the mean trajectory and the other based on being accurate in high probability (IHP). An off-line optimization problem is solved to find the constant step size and number of iterations to achieve the desired tracking accuracy. Simulations are used to confirm that this approach provides the desired tracking accuracy.

Original language	English (US)
Article number	7039377
Pages (from-to)	173-178
Number of pages	6
Journal	Proceedings of the IEEE Conference on Decision and Control
Volume	2015-February
Issue number	February
DOIs	https://doi.org/10.1109/CDC.2014.7039377
State	Published - 2014
Externally published	Yes
Event	2014 53rd IEEE Annual Conference on Decision and Control, CDC 2014 - Los Angeles, United States Duration: Dec 15 2014 → Dec 17 2014

Keywords

adaptive optimization
gradient methods
stochastic optimization

ASJC Scopus subject areas

Control and Systems Engineering
Modeling and Simulation
Control and Optimization

Access to Document

10.1109/CDC.2014.7039377

Cite this

@article{2ae4640e00284c91ae88c84e2ede665f,

title = "Dynamic stochastic optimization",

abstract = "A framework for sequentially solving stochastic optimization problems with stochastic gradient descent is introduced. Two tracking criteria are considered, one based on being accurate with respect to the mean trajectory and the other based on being accurate in high probability (IHP). An off-line optimization problem is solved to find the constant step size and number of iterations to achieve the desired tracking accuracy. Simulations are used to confirm that this approach provides the desired tracking accuracy.",

keywords = "adaptive optimization, gradient methods, stochastic optimization",

author = "Craig Wilson and Venugopal Veeravalli and Angelia Nedic",

note = "Publisher Copyright: {\textcopyright} 2014 IEEE.; 2014 53rd IEEE Annual Conference on Decision and Control, CDC 2014 ; Conference date: 15-12-2014 Through 17-12-2014",

year = "2014",

doi = "10.1109/CDC.2014.7039377",

language = "English (US)",

volume = "2015-February",

pages = "173--178",

journal = "Proceedings of the IEEE Conference on Decision and Control",

issn = "0743-1546",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

number = "February",

}

TY - JOUR

T1 - Dynamic stochastic optimization

AU - Wilson, Craig

AU - Veeravalli, Venugopal

AU - Nedic, Angelia

PY - 2014

Y1 - 2014

N2 - A framework for sequentially solving stochastic optimization problems with stochastic gradient descent is introduced. Two tracking criteria are considered, one based on being accurate with respect to the mean trajectory and the other based on being accurate in high probability (IHP). An off-line optimization problem is solved to find the constant step size and number of iterations to achieve the desired tracking accuracy. Simulations are used to confirm that this approach provides the desired tracking accuracy.

AB - A framework for sequentially solving stochastic optimization problems with stochastic gradient descent is introduced. Two tracking criteria are considered, one based on being accurate with respect to the mean trajectory and the other based on being accurate in high probability (IHP). An off-line optimization problem is solved to find the constant step size and number of iterations to achieve the desired tracking accuracy. Simulations are used to confirm that this approach provides the desired tracking accuracy.

KW - adaptive optimization

KW - gradient methods

KW - stochastic optimization

UR - http://www.scopus.com/inward/record.url?scp=84988286555&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84988286555&partnerID=8YFLogxK

U2 - 10.1109/CDC.2014.7039377

DO - 10.1109/CDC.2014.7039377

M3 - Conference article

AN - SCOPUS:84988286555

SN - 0743-1546

VL - 2015-February

SP - 173

EP - 178

JO - Proceedings of the IEEE Conference on Decision and Control

JF - Proceedings of the IEEE Conference on Decision and Control

IS - February

M1 - 7039377

T2 - 2014 53rd IEEE Annual Conference on Decision and Control, CDC 2014

Y2 - 15 December 2014 through 17 December 2014

ER -

Dynamic stochastic optimization

Abstract

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this