Multivariate exponentially weighted moving covariance matrix

Douglas M. Hawkins; Edgard M. Maboudou-Tchao

doi:10.1198/004017008000000163

Multivariate exponentially weighted moving covariance matrix

Douglas M. Hawkins, Edgard M. Maboudou-Tchao

Statistics (Twin Cities)

Research output: Contribution to journal › Article › peer-review

105 Scopus citations

Abstract

Multivariate exponentially weighted moving average (MEWMA) charts are among the best control charts for detecting small changes in any direction. The well-known MEWMA is directed at changes in the mean vector. But changes can occur in either the location or the variability of the correlated multivariate quality characteristics, calling for parallel methodologies for detecting changes in the covariance matrix. This article discusses an exponentially weighted moving covariance matrix for monitoring the stability of the covariance matrix of a process. Used together with the location MEWMA, this chart provides a way to satisfy Shewhart's dictum that proper process control monitor both mean and variability. We show that the chart is competitive, generally outperforming current control charts for the covariance matrix.

Original language	English (US)
Pages (from-to)	155-166
Number of pages	12
Journal	Technometrics
Volume	50
Issue number	2
DOIs	https://doi.org/10.1198/004017008000000163
State	Published - May 2008

Bibliographical note

Funding Information:
This work was supported by the National Science Foundation grant DMS 0306304. The authors are grateful to the editors and referee for a number of helpful suggestions.

Keywords

Average run length
Average run length bias
Regression adjustment

Publisher link

10.1198/004017008000000163

Find It

Find It at U of M Libraries

Cite this

@article{02d502a90247454e94c2bf0bad5c61ca,

title = "Multivariate exponentially weighted moving covariance matrix",

abstract = "Multivariate exponentially weighted moving average (MEWMA) charts are among the best control charts for detecting small changes in any direction. The well-known MEWMA is directed at changes in the mean vector. But changes can occur in either the location or the variability of the correlated multivariate quality characteristics, calling for parallel methodologies for detecting changes in the covariance matrix. This article discusses an exponentially weighted moving covariance matrix for monitoring the stability of the covariance matrix of a process. Used together with the location MEWMA, this chart provides a way to satisfy Shewhart's dictum that proper process control monitor both mean and variability. We show that the chart is competitive, generally outperforming current control charts for the covariance matrix.",

keywords = "Average run length, Average run length bias, Regression adjustment",

author = "Hawkins, {Douglas M.} and Maboudou-Tchao, {Edgard M.}",

note = "Funding Information: This work was supported by the National Science Foundation grant DMS 0306304. The authors are grateful to the editors and referee for a number of helpful suggestions.",

year = "2008",

month = may,

doi = "10.1198/004017008000000163",

language = "English (US)",

volume = "50",

pages = "155--166",

journal = "Technometrics",

issn = "0040-1706",

publisher = "Taylor and Francis Ltd.",

number = "2",

}

TY - JOUR

T1 - Multivariate exponentially weighted moving covariance matrix

AU - Hawkins, Douglas M.

AU - Maboudou-Tchao, Edgard M.

N1 - Funding Information: This work was supported by the National Science Foundation grant DMS 0306304. The authors are grateful to the editors and referee for a number of helpful suggestions.

PY - 2008/5

Y1 - 2008/5

N2 - Multivariate exponentially weighted moving average (MEWMA) charts are among the best control charts for detecting small changes in any direction. The well-known MEWMA is directed at changes in the mean vector. But changes can occur in either the location or the variability of the correlated multivariate quality characteristics, calling for parallel methodologies for detecting changes in the covariance matrix. This article discusses an exponentially weighted moving covariance matrix for monitoring the stability of the covariance matrix of a process. Used together with the location MEWMA, this chart provides a way to satisfy Shewhart's dictum that proper process control monitor both mean and variability. We show that the chart is competitive, generally outperforming current control charts for the covariance matrix.

AB - Multivariate exponentially weighted moving average (MEWMA) charts are among the best control charts for detecting small changes in any direction. The well-known MEWMA is directed at changes in the mean vector. But changes can occur in either the location or the variability of the correlated multivariate quality characteristics, calling for parallel methodologies for detecting changes in the covariance matrix. This article discusses an exponentially weighted moving covariance matrix for monitoring the stability of the covariance matrix of a process. Used together with the location MEWMA, this chart provides a way to satisfy Shewhart's dictum that proper process control monitor both mean and variability. We show that the chart is competitive, generally outperforming current control charts for the covariance matrix.

KW - Average run length

KW - Average run length bias

KW - Regression adjustment

UR - http://www.scopus.com/inward/record.url?scp=46749103696&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=46749103696&partnerID=8YFLogxK

U2 - 10.1198/004017008000000163

DO - 10.1198/004017008000000163

M3 - Article

AN - SCOPUS:46749103696

SN - 0040-1706

VL - 50

SP - 155

EP - 166

JO - Technometrics

JF - Technometrics

IS - 2

ER -

Multivariate exponentially weighted moving covariance matrix

Abstract

Bibliographical note

Keywords

Publisher link

Find It

Other files and links

Fingerprint

Cite this