Network independent rates in distributed learning

Angelia Nedić; Alex Olshevsky; César A. Uribe

doi:10.1109/ACC.2016.7525057

Network independent rates in distributed learning

Angelia Nedić, Alex Olshevsky, César A. Uribe

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

18 Scopus citations

Abstract

We propose a novel belief update algorithm for Distributed Non-Bayesian learning over time-varying directed graphs, where a group of agents tries to collectively select a distribution that best describes the observed data. We show that the proposed update rule, inspired by the Push-Sum algorithm, is consistent; moreover we provide an explicit characterization of its convergence rate. Our main result states that, after a transient time, all agents will concentrate their beliefs at a network independent rate. Network independent rates were not available for other consensus based distributed learning algorithms on time-varying directed graphs.

Original language	English (US)
Title of host publication	2016 American Control Conference, ACC 2016
Publisher	Institute of Electrical and Electronics Engineers Inc.
Pages	1072-1077
Number of pages	6
ISBN (Electronic)	9781467386821
DOIs	https://doi.org/10.1109/ACC.2016.7525057
State	Published - Jul 28 2016
Externally published	Yes
Event	2016 American Control Conference, ACC 2016 - Boston, United States Duration: Jul 6 2016 → Jul 8 2016

Publication series

Name	Proceedings of the American Control Conference
Volume	2016-July
ISSN (Print)	0743-1619

Other

Other	2016 American Control Conference, ACC 2016
Country/Territory	United States
City	Boston
Period	7/6/16 → 7/8/16

ASJC Scopus subject areas

Electrical and Electronic Engineering

Access to Document

10.1109/ACC.2016.7525057

Cite this

Network independent rates in distributed learning. / Nedić, Angelia; Olshevsky, Alex; Uribe, César A.
2016 American Control Conference, ACC 2016. Institute of Electrical and Electronics Engineers Inc., 2016. p. 1072-1077 7525057 (Proceedings of the American Control Conference; Vol. 2016-July).

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

Nedić, A, Olshevsky, A & Uribe, CA 2016, Network independent rates in distributed learning. in 2016 American Control Conference, ACC 2016., 7525057, Proceedings of the American Control Conference, vol. 2016-July, Institute of Electrical and Electronics Engineers Inc., pp. 1072-1077, 2016 American Control Conference, ACC 2016, Boston, United States, 7/6/16. https://doi.org/10.1109/ACC.2016.7525057

@inproceedings{3146011ceecd4dad82e62d8a8b979dd8,

title = "Network independent rates in distributed learning",

abstract = "We propose a novel belief update algorithm for Distributed Non-Bayesian learning over time-varying directed graphs, where a group of agents tries to collectively select a distribution that best describes the observed data. We show that the proposed update rule, inspired by the Push-Sum algorithm, is consistent; moreover we provide an explicit characterization of its convergence rate. Our main result states that, after a transient time, all agents will concentrate their beliefs at a network independent rate. Network independent rates were not available for other consensus based distributed learning algorithms on time-varying directed graphs.",

author = "Angelia Nedi{\'c} and Alex Olshevsky and Uribe, {C{\'e}sar A.}",

note = "Publisher Copyright: {\textcopyright} 2016 American Automatic Control Council (AACC).; 2016 American Control Conference, ACC 2016 ; Conference date: 06-07-2016 Through 08-07-2016",

year = "2016",

month = jul,

day = "28",

doi = "10.1109/ACC.2016.7525057",

language = "English (US)",

series = "Proceedings of the American Control Conference",

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

pages = "1072--1077",

booktitle = "2016 American Control Conference, ACC 2016",

}

TY - GEN

T1 - Network independent rates in distributed learning

AU - Nedić, Angelia

AU - Olshevsky, Alex

AU - Uribe, César A.

PY - 2016/7/28

Y1 - 2016/7/28

N2 - We propose a novel belief update algorithm for Distributed Non-Bayesian learning over time-varying directed graphs, where a group of agents tries to collectively select a distribution that best describes the observed data. We show that the proposed update rule, inspired by the Push-Sum algorithm, is consistent; moreover we provide an explicit characterization of its convergence rate. Our main result states that, after a transient time, all agents will concentrate their beliefs at a network independent rate. Network independent rates were not available for other consensus based distributed learning algorithms on time-varying directed graphs.

AB - We propose a novel belief update algorithm for Distributed Non-Bayesian learning over time-varying directed graphs, where a group of agents tries to collectively select a distribution that best describes the observed data. We show that the proposed update rule, inspired by the Push-Sum algorithm, is consistent; moreover we provide an explicit characterization of its convergence rate. Our main result states that, after a transient time, all agents will concentrate their beliefs at a network independent rate. Network independent rates were not available for other consensus based distributed learning algorithms on time-varying directed graphs.

UR - http://www.scopus.com/inward/record.url?scp=84992053056&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84992053056&partnerID=8YFLogxK

U2 - 10.1109/ACC.2016.7525057

DO - 10.1109/ACC.2016.7525057

M3 - Conference contribution

AN - SCOPUS:84992053056

T3 - Proceedings of the American Control Conference

SP - 1072

EP - 1077

BT - 2016 American Control Conference, ACC 2016

PB - Institute of Electrical and Electronics Engineers Inc.

T2 - 2016 American Control Conference, ACC 2016

Y2 - 6 July 2016 through 8 July 2016

ER -

Network independent rates in distributed learning

Abstract

Publication series

Other

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this