Please use this identifier to cite or link to this item:
https://open.uns.ac.rs/handle/123456789/585
DC Field | Value | Language |
---|---|---|
dc.contributor.author | Jakovetić, Dušan | en |
dc.contributor.author | Bajović, Dragana | en |
dc.contributor.author | Sahu A. | en |
dc.contributor.author | Kar S. | en |
dc.date.accessioned | 2019-09-23T10:09:10Z | - |
dc.date.available | 2019-09-23T10:09:10Z | - |
dc.date.issued | 2019-01-18 | en |
dc.identifier.isbn | 9781538613955 | en |
dc.identifier.issn | 07431546 | en |
dc.identifier.uri | https://open.uns.ac.rs/handle/123456789/585 | - |
dc.description.abstract | © 2018 IEEE. We establish the O(\frac{1}{k}) convergence rate for distributed stochastic gradient methods that operate over strongly convex costs and random networks. The considered class of methods is standard - each node performs a weighted average of its own and its neighbors' solution estimates (consensus), and takes a negative step with respect to a noisy version of its local function's gradient (innovation). The underlying communication network is modeled through a sequence of temporally independent identically distributed (i.i.d.) Laplacian matrices such that the underlying graphs are connected on average; the local gradient noises are also i.i.d. in time, have finite second moment, and possibly unbounded support. We show that, after a careful setting of the consensus and innovations potentials (weights), the distributed stochastic gradient method achieves a (order-optimal) O(\frac{1}{k}) convergence rate in the mean square distance from the solution. To the best of our knowledge, this is the first order-optimal convergence rate result on distributed strongly convex stochastic optimization when the network is random and the gradient noises have unbounded support. Simulation examples confirm the theoretical findings. | en |
dc.relation.ispartof | Proceedings of the IEEE Conference on Decision and Control | en |
dc.title | Convergence Rates for Distributed Stochastic Optimization over Random Networks | en |
dc.type | Conference Paper | en |
dc.identifier.doi | 10.1109/CDC.2018.8619228 | en |
dc.identifier.scopus | 2-s2.0-85062172966 | en |
dc.identifier.url | https://api.elsevier.com/content/abstract/scopus_id/85062172966 | en |
dc.relation.lastpage | 4245 | en |
dc.relation.firstpage | 4238 | en |
dc.relation.volume | 2018-December | en |
item.grantfulltext | none | - |
item.fulltext | No Fulltext | - |
crisitem.author.dept | Prirodno-matematički fakultet, Departman za matematiku i informatiku | - |
crisitem.author.dept | Fakultet tehničkih nauka, Departman za energetiku, elektroniku i telekomunikacije | - |
crisitem.author.parentorg | Prirodno-matematički fakultet | - |
crisitem.author.parentorg | Fakultet tehničkih nauka | - |
Appears in Collections: | FTN Publikacije/Publications |
SCOPUSTM
Citations
34
checked on May 10, 2024
Page view(s)
15
Last Week
5
5
Last month
0
0
checked on May 10, 2024
Google ScholarTM
Check
Altmetric
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.