Open Collections

UBC Faculty Research and Publications

Convergence rate of stochastic gradient with constant step size Schmidt, Mark

Abstract

We show that the basic stochastic gradient method applied to a strongly-convex differentiable function with a constant step-size achieves a linear convergence rate (in function value and iterates) up to a constant proportional the step-size (under standard assumptions on the gradient).

Item Metadata

Title
Convergence rate of stochastic gradient with constant step size
Creator
Schmidt, Mark
Date Issued
2014-09-05
Description
We show that the basic stochastic gradient method applied to a strongly-convex differentiable function with a constant step-size achieves a linear convergence rate (in function value and iterates) up to a constant proportional the step-size (under standard assumptions on the gradient).
Genre
Article
Type
Text
Language
eng
Date Available
2014-09-15
Provider
Vancouver : University of British Columbia Library
Rights
Attribution-NoDerivs 2.5 Canada
DOI
10.14288/1.0050992
URI
http://hdl.handle.net/2429/50358
Affiliation
Science, Faculty of; Computer Science, Department of
Peer Review Status
Unreviewed
Scholarly Level
Faculty
Rights URI
http://creativecommons.org/licenses/by-nd/2.5/ca/
Aggregated Source Repository
DSpace

Item Media

Item Citations and Data

Rights

Attribution-NoDerivs 2.5 Canada