Distributed random projection algorithm for convex optimization

Soomin Lee, Angelia Nedich

Research output: Contribution to journalArticle

102 Citations (Scopus)

Abstract

Random projection algorithm is of interest for constrained optimization when the constraint set is not known in advance or the projection operation on the whole constraint set is computationally prohibitive. This paper presents a distributed random projection algorithm for constrained convex optimization problems that can be used by multiple agents connected over a time-varying network, where each agent has its own objective function and its own constrained set. We prove that the iterates of all agents converge to the same point in the optimal set almost surely. Experiments on distributed support vector machines demonstrate good performance of the algorithm.

Original languageEnglish (US)
Article number6461383
Pages (from-to)221-229
Number of pages9
JournalIEEE Journal on Selected Topics in Signal Processing
Volume7
Issue number2
DOIs
StatePublished - 2013
Externally publishedYes

Fingerprint

Convex optimization
Constrained optimization
Time varying networks
Support vector machines
Experiments

Keywords

  • Asynchronous algorithms
  • Distributed convex optimization
  • Distributed multi-agent system
  • Random gossip network

ASJC Scopus subject areas

  • Electrical and Electronic Engineering
  • Signal Processing

Cite this

Distributed random projection algorithm for convex optimization. / Lee, Soomin; Nedich, Angelia.

In: IEEE Journal on Selected Topics in Signal Processing, Vol. 7, No. 2, 6461383, 2013, p. 221-229.

Research output: Contribution to journalArticle

@article{a53e24ec95f143dfb6102b6e2bba67bd,
title = "Distributed random projection algorithm for convex optimization",
abstract = "Random projection algorithm is of interest for constrained optimization when the constraint set is not known in advance or the projection operation on the whole constraint set is computationally prohibitive. This paper presents a distributed random projection algorithm for constrained convex optimization problems that can be used by multiple agents connected over a time-varying network, where each agent has its own objective function and its own constrained set. We prove that the iterates of all agents converge to the same point in the optimal set almost surely. Experiments on distributed support vector machines demonstrate good performance of the algorithm.",
keywords = "Asynchronous algorithms, Distributed convex optimization, Distributed multi-agent system, Random gossip network",
author = "Soomin Lee and Angelia Nedich",
year = "2013",
doi = "10.1109/JSTSP.2013.2247023",
language = "English (US)",
volume = "7",
pages = "221--229",
journal = "IEEE Journal on Selected Topics in Signal Processing",
issn = "1932-4553",
publisher = "Institute of Electrical and Electronics Engineers Inc.",
number = "2",

}

TY - JOUR

T1 - Distributed random projection algorithm for convex optimization

AU - Lee, Soomin

AU - Nedich, Angelia

PY - 2013

Y1 - 2013

N2 - Random projection algorithm is of interest for constrained optimization when the constraint set is not known in advance or the projection operation on the whole constraint set is computationally prohibitive. This paper presents a distributed random projection algorithm for constrained convex optimization problems that can be used by multiple agents connected over a time-varying network, where each agent has its own objective function and its own constrained set. We prove that the iterates of all agents converge to the same point in the optimal set almost surely. Experiments on distributed support vector machines demonstrate good performance of the algorithm.

AB - Random projection algorithm is of interest for constrained optimization when the constraint set is not known in advance or the projection operation on the whole constraint set is computationally prohibitive. This paper presents a distributed random projection algorithm for constrained convex optimization problems that can be used by multiple agents connected over a time-varying network, where each agent has its own objective function and its own constrained set. We prove that the iterates of all agents converge to the same point in the optimal set almost surely. Experiments on distributed support vector machines demonstrate good performance of the algorithm.

KW - Asynchronous algorithms

KW - Distributed convex optimization

KW - Distributed multi-agent system

KW - Random gossip network

UR - http://www.scopus.com/inward/record.url?scp=84890517050&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84890517050&partnerID=8YFLogxK

U2 - 10.1109/JSTSP.2013.2247023

DO - 10.1109/JSTSP.2013.2247023

M3 - Article

AN - SCOPUS:84890517050

VL - 7

SP - 221

EP - 229

JO - IEEE Journal on Selected Topics in Signal Processing

JF - IEEE Journal on Selected Topics in Signal Processing

SN - 1932-4553

IS - 2

M1 - 6461383

ER -