Tracking performance of online stochastic learners
File(s)
Accepted version
Author(s)
Vlaski, Stefan
Rizk, Elsa
Sayed, Ali H
Type
Journal Article
Abstract
The utilization of online stochastic algorithms is popular in large-scale learning settings due to their ability to compute updates on the fly, without the need to store and process data in large batches. When a constant step-size is used, these algorithms also have the ability to adapt to drifts in problem parameters, such as data or model properties, and track the optimal solution with reasonable accuracy. Building on analogies with the study of adaptive filters, we establish a link between steady-state performance derived under stationarity assumptions and the tracking performance of online learners under random walk models. The link allows us to infer the tracking performance from steady-state expressions directly and almost by inspection.
Date Issued
2020
Date Acceptance
2020-07-21
Citation
IEEE Signal Processing Letters, 2020, 27, pp.1385-1389
ISSN
1070-9908
Publisher
Institute of Electrical and Electronics Engineers
Start Page
1385
End Page
1389
Journal / Book Title
IEEE Signal Processing Letters
Volume
27
Copyright Statement
Copyright © 2020 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works.
Identifier
https://www.webofscience.com/api/gateway?GWVersion=2&SrcApp=PARTNER_APP&SrcAuth=LinksAMR&KeyUT=WOS:000562025400003&DestLinkType=FullRecord&DestApp=ALL_WOS&UsrCustomerID=a2bf6146997ec60c407a63945d4e92bb
Subjects
Adaptation models
Engineering
Engineering, Electrical & Electronic
GRAPHS
Noise measurement
non-stationary environment
Online learning
Optimization
OPTIMIZATION
Random variables
Science & Technology
Signal processing algorithms
Steady-state
stochastic learning
Stochastic processes
Technology
tracking performance
Publication Status
Published
Date Publish Online
2020-08-03