Distributed optimization of strongly convex functions on directed time-varying graphs

Angelia Nedic; Alex Olshevsky

doi:10.1109/GlobalSIP.2013.6736882

Distributed optimization of strongly convex functions on directed time-varying graphs

Angelia Nedic, Alex Olshevsky

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

10 Scopus citations

Abstract

We investigate the convergence rate of the recently proposed subgadient-push method for distributed separable optimization over time-varying directed graphs. The algorithm requires no knowledge of either the number of agents or the graph sequence to implement, nor the use of doubly stochastic weights. We show that the algorithm converges at a rate of O(ln t/t) for strongly convex functions. The proportionality constant in the rate estimate depends on some problem parameters, the initial values at the nodes, the speed of the network information diffusion, and the imbalances of influence among the nodes.

Original language	English (US)
Title of host publication	2013 IEEE Global Conference on Signal and Information Processing, GlobalSIP 2013 - Proceedings
Pages	329-332
Number of pages	4
DOIs	https://doi.org/10.1109/GlobalSIP.2013.6736882
State	Published - Dec 1 2013
Externally published	Yes
Event	2013 1st IEEE Global Conference on Signal and Information Processing, GlobalSIP 2013 - Austin, TX, United States Duration: Dec 3 2013 → Dec 5 2013

Publication series

Name	2013 IEEE Global Conference on Signal and Information Processing, GlobalSIP 2013 - Proceedings

Other

Other	2013 1st IEEE Global Conference on Signal and Information Processing, GlobalSIP 2013
Country/Territory	United States
City	Austin, TX
Period	12/3/13 → 12/5/13

ASJC Scopus subject areas

Information Systems
Signal Processing

Access to Document

10.1109/GlobalSIP.2013.6736882

Cite this

Nedic, A., & Olshevsky, A. (2013). Distributed optimization of strongly convex functions on directed time-varying graphs. In 2013 IEEE Global Conference on Signal and Information Processing, GlobalSIP 2013 - Proceedings (pp. 329-332). Article 6736882 (2013 IEEE Global Conference on Signal and Information Processing, GlobalSIP 2013 - Proceedings). https://doi.org/10.1109/GlobalSIP.2013.6736882

Distributed optimization of strongly convex functions on directed time-varying graphs. / Nedic, Angelia; Olshevsky, Alex.
2013 IEEE Global Conference on Signal and Information Processing, GlobalSIP 2013 - Proceedings. 2013. p. 329-332 6736882 (2013 IEEE Global Conference on Signal and Information Processing, GlobalSIP 2013 - Proceedings).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Nedic, A & Olshevsky, A 2013, Distributed optimization of strongly convex functions on directed time-varying graphs. in 2013 IEEE Global Conference on Signal and Information Processing, GlobalSIP 2013 - Proceedings., 6736882, 2013 IEEE Global Conference on Signal and Information Processing, GlobalSIP 2013 - Proceedings, pp. 329-332, 2013 1st IEEE Global Conference on Signal and Information Processing, GlobalSIP 2013, Austin, TX, United States, 12/3/13. https://doi.org/10.1109/GlobalSIP.2013.6736882

@inproceedings{3be739cdec1447a5b02a174111b6c879,

title = "Distributed optimization of strongly convex functions on directed time-varying graphs",

abstract = "We investigate the convergence rate of the recently proposed subgadient-push method for distributed separable optimization over time-varying directed graphs. The algorithm requires no knowledge of either the number of agents or the graph sequence to implement, nor the use of doubly stochastic weights. We show that the algorithm converges at a rate of O(ln t/t) for strongly convex functions. The proportionality constant in the rate estimate depends on some problem parameters, the initial values at the nodes, the speed of the network information diffusion, and the imbalances of influence among the nodes.",

author = "Angelia Nedic and Alex Olshevsky",

year = "2013",

month = dec,

day = "1",

doi = "10.1109/GlobalSIP.2013.6736882",

language = "English (US)",

isbn = "9781479902484",

series = "2013 IEEE Global Conference on Signal and Information Processing, GlobalSIP 2013 - Proceedings",

pages = "329--332",

booktitle = "2013 IEEE Global Conference on Signal and Information Processing, GlobalSIP 2013 - Proceedings",

note = "2013 1st IEEE Global Conference on Signal and Information Processing, GlobalSIP 2013 ; Conference date: 03-12-2013 Through 05-12-2013",

}

TY - GEN

T1 - Distributed optimization of strongly convex functions on directed time-varying graphs

AU - Nedic, Angelia

AU - Olshevsky, Alex

PY - 2013/12/1

Y1 - 2013/12/1

N2 - We investigate the convergence rate of the recently proposed subgadient-push method for distributed separable optimization over time-varying directed graphs. The algorithm requires no knowledge of either the number of agents or the graph sequence to implement, nor the use of doubly stochastic weights. We show that the algorithm converges at a rate of O(ln t/t) for strongly convex functions. The proportionality constant in the rate estimate depends on some problem parameters, the initial values at the nodes, the speed of the network information diffusion, and the imbalances of influence among the nodes.

AB - We investigate the convergence rate of the recently proposed subgadient-push method for distributed separable optimization over time-varying directed graphs. The algorithm requires no knowledge of either the number of agents or the graph sequence to implement, nor the use of doubly stochastic weights. We show that the algorithm converges at a rate of O(ln t/t) for strongly convex functions. The proportionality constant in the rate estimate depends on some problem parameters, the initial values at the nodes, the speed of the network information diffusion, and the imbalances of influence among the nodes.

UR - http://www.scopus.com/inward/record.url?scp=84897687535&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84897687535&partnerID=8YFLogxK

U2 - 10.1109/GlobalSIP.2013.6736882

DO - 10.1109/GlobalSIP.2013.6736882

M3 - Conference contribution

AN - SCOPUS:84897687535

SN - 9781479902484

T3 - 2013 IEEE Global Conference on Signal and Information Processing, GlobalSIP 2013 - Proceedings

SP - 329

EP - 332

BT - 2013 IEEE Global Conference on Signal and Information Processing, GlobalSIP 2013 - Proceedings

T2 - 2013 1st IEEE Global Conference on Signal and Information Processing, GlobalSIP 2013

Y2 - 3 December 2013 through 5 December 2013

ER -

Distributed optimization of strongly convex functions on directed time-varying graphs

Abstract

Publication series

Other

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this