One-class classification-based control charts for multivariate process monitoring

Thuntee Sukchotrat, Seoung Bum Kim, Fugee Tsung

Research output: Contribution to journalArticle

58 Citations (Scopus)

Abstract

One-class classification problems have attracted a great deal of attention from various disciplines. In the present study, attempts are made to extend the scope of application of the one-class classification technique to Statistical Process Control (SPC) problems. New multivariate control charts that apply the effectiveness of one-class classification to improvement of Phase I and Phase II analysis in SPC are proposed. These charts use a monitoring statistic to represent the degree of being an outlier as obtained through one-class classification. The control limits of the proposed charts are established based on the empirical level of significance on the percentile, estimated by the bootstrap method. A simulation study is conducted to illustrate the limitations of current one-class classification control charts and demonstrate the effectiveness of the proposed control charts.

Original languageEnglish
Pages (from-to)107-120
Number of pages14
JournalIIE Transactions (Institute of Industrial Engineers)
Volume42
Issue number2
DOIs
Publication statusPublished - 2010 Feb 1

Fingerprint

Process monitoring
Statistical process control
Control charts
Statistics
Monitoring

Keywords

  • Data mining
  • Hotelling's T
  • Multivariate process
  • One-class classification method
  • Statistical process control

ASJC Scopus subject areas

  • Industrial and Manufacturing Engineering

Cite this

One-class classification-based control charts for multivariate process monitoring. / Sukchotrat, Thuntee; Kim, Seoung Bum; Tsung, Fugee.

In: IIE Transactions (Institute of Industrial Engineers), Vol. 42, No. 2, 01.02.2010, p. 107-120.

Research output: Contribution to journalArticle

@article{d993964df123417eaa1bbb19f965e5e2,
title = "One-class classification-based control charts for multivariate process monitoring",
abstract = "One-class classification problems have attracted a great deal of attention from various disciplines. In the present study, attempts are made to extend the scope of application of the one-class classification technique to Statistical Process Control (SPC) problems. New multivariate control charts that apply the effectiveness of one-class classification to improvement of Phase I and Phase II analysis in SPC are proposed. These charts use a monitoring statistic to represent the degree of being an outlier as obtained through one-class classification. The control limits of the proposed charts are established based on the empirical level of significance on the percentile, estimated by the bootstrap method. A simulation study is conducted to illustrate the limitations of current one-class classification control charts and demonstrate the effectiveness of the proposed control charts.",
keywords = "Data mining, Hotelling's T, Multivariate process, One-class classification method, Statistical process control",
author = "Thuntee Sukchotrat and Kim, {Seoung Bum} and Fugee Tsung",
year = "2010",
month = "2",
day = "1",
doi = "10.1080/07408170903019150",
language = "English",
volume = "42",
pages = "107--120",
journal = "IISE Transactions",
issn = "2472-5854",
publisher = "Taylor and Francis Ltd.",
number = "2",

}

TY - JOUR

T1 - One-class classification-based control charts for multivariate process monitoring

AU - Sukchotrat, Thuntee

AU - Kim, Seoung Bum

AU - Tsung, Fugee

PY - 2010/2/1

Y1 - 2010/2/1

N2 - One-class classification problems have attracted a great deal of attention from various disciplines. In the present study, attempts are made to extend the scope of application of the one-class classification technique to Statistical Process Control (SPC) problems. New multivariate control charts that apply the effectiveness of one-class classification to improvement of Phase I and Phase II analysis in SPC are proposed. These charts use a monitoring statistic to represent the degree of being an outlier as obtained through one-class classification. The control limits of the proposed charts are established based on the empirical level of significance on the percentile, estimated by the bootstrap method. A simulation study is conducted to illustrate the limitations of current one-class classification control charts and demonstrate the effectiveness of the proposed control charts.

AB - One-class classification problems have attracted a great deal of attention from various disciplines. In the present study, attempts are made to extend the scope of application of the one-class classification technique to Statistical Process Control (SPC) problems. New multivariate control charts that apply the effectiveness of one-class classification to improvement of Phase I and Phase II analysis in SPC are proposed. These charts use a monitoring statistic to represent the degree of being an outlier as obtained through one-class classification. The control limits of the proposed charts are established based on the empirical level of significance on the percentile, estimated by the bootstrap method. A simulation study is conducted to illustrate the limitations of current one-class classification control charts and demonstrate the effectiveness of the proposed control charts.

KW - Data mining

KW - Hotelling's T

KW - Multivariate process

KW - One-class classification method

KW - Statistical process control

UR - http://www.scopus.com/inward/record.url?scp=77649328220&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=77649328220&partnerID=8YFLogxK

U2 - 10.1080/07408170903019150

DO - 10.1080/07408170903019150

M3 - Article

VL - 42

SP - 107

EP - 120

JO - IISE Transactions

JF - IISE Transactions

SN - 2472-5854

IS - 2

ER -