Weighted averaging and stochastic approximation

I. J. Wang, Edwin K.P. Chong, Sanjeev R. Kulkarni

Research output: Contribution to journalArticle

18 Citations (Scopus)

Abstract

We explore the relationship between weighted averaging and stochastic approximation algorithms, and study their convergence via a sample-path analysis. We prove that the convergence of a stochastic approximation algorithm is equivalent to the convergence of the weighted average of the associated noise sequence. We also present necessary and sufficient noise conditions for convergence of the average of the output of a stochastic approximation algorithm in the linear case. We show that the averaged stochastic approximation algorithms can tolerate a larger class of noise sequences than the stand-alone stochastic approximation algorithms.

Original languageEnglish (US)
Pages (from-to)41-60
Number of pages20
JournalMathematics of Control, Signals, and Systems
Volume10
Issue number1
DOIs
StatePublished - Jan 1 1997

Fingerprint

Stochastic Approximation
Stochastic Algorithms
Approximation algorithms
Averaging
Approximation Algorithms
Path Analysis
Sample Path
Weighted Average
Sufficient
Necessary
Output

All Science Journal Classification (ASJC) codes

  • Control and Optimization
  • Signal Processing
  • Applied Mathematics
  • Control and Systems Engineering

Cite this

@article{35b0a6a972d243238278c7371acf9267,
title = "Weighted averaging and stochastic approximation",
abstract = "We explore the relationship between weighted averaging and stochastic approximation algorithms, and study their convergence via a sample-path analysis. We prove that the convergence of a stochastic approximation algorithm is equivalent to the convergence of the weighted average of the associated noise sequence. We also present necessary and sufficient noise conditions for convergence of the average of the output of a stochastic approximation algorithm in the linear case. We show that the averaged stochastic approximation algorithms can tolerate a larger class of noise sequences than the stand-alone stochastic approximation algorithms.",
author = "Wang, {I. J.} and Chong, {Edwin K.P.} and Kulkarni, {Sanjeev R.}",
year = "1997",
month = "1",
day = "1",
doi = "https://doi.org/10.1007/BF01219775",
language = "English (US)",
volume = "10",
pages = "41--60",
journal = "Mathematics of Control, Signals, and Systems",
issn = "0932-4194",
publisher = "Springer London",
number = "1",

}

Weighted averaging and stochastic approximation. / Wang, I. J.; Chong, Edwin K.P.; Kulkarni, Sanjeev R.

In: Mathematics of Control, Signals, and Systems, Vol. 10, No. 1, 01.01.1997, p. 41-60.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Weighted averaging and stochastic approximation

AU - Wang, I. J.

AU - Chong, Edwin K.P.

AU - Kulkarni, Sanjeev R.

PY - 1997/1/1

Y1 - 1997/1/1

N2 - We explore the relationship between weighted averaging and stochastic approximation algorithms, and study their convergence via a sample-path analysis. We prove that the convergence of a stochastic approximation algorithm is equivalent to the convergence of the weighted average of the associated noise sequence. We also present necessary and sufficient noise conditions for convergence of the average of the output of a stochastic approximation algorithm in the linear case. We show that the averaged stochastic approximation algorithms can tolerate a larger class of noise sequences than the stand-alone stochastic approximation algorithms.

AB - We explore the relationship between weighted averaging and stochastic approximation algorithms, and study their convergence via a sample-path analysis. We prove that the convergence of a stochastic approximation algorithm is equivalent to the convergence of the weighted average of the associated noise sequence. We also present necessary and sufficient noise conditions for convergence of the average of the output of a stochastic approximation algorithm in the linear case. We show that the averaged stochastic approximation algorithms can tolerate a larger class of noise sequences than the stand-alone stochastic approximation algorithms.

UR - http://www.scopus.com/inward/record.url?scp=0030658716&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=0030658716&partnerID=8YFLogxK

U2 - https://doi.org/10.1007/BF01219775

DO - https://doi.org/10.1007/BF01219775

M3 - Article

VL - 10

SP - 41

EP - 60

JO - Mathematics of Control, Signals, and Systems

JF - Mathematics of Control, Signals, and Systems

SN - 0932-4194

IS - 1

ER -