Integration of support vector machines and control charts for multivariate process monitoring

Panitarn Chongfuangprinya, Seoung Bum Kim, Sun Kyoung Park, Thuntee Sukchotrat

Research output: Contribution to journalArticle

13 Citations (Scopus)

Abstract

Statistical process control tools have been used routinely to improve process capabilities through reliable on-line monitoring and diagnostic processes. In the present paper, we propose a novel multivariate control chart that integrates a support vector machine (SVM) algorithm, a bootstrap method, and a control chart technique to improve multivariate process monitoring. The proposed chart uses as the monitoring statistic the predicted probability of class (PoC) values from an SVM algorithm. The control limits of SVM-PoC charts are obtained by a bootstrap approach.A simulation study was conducted to evaluate the performance of the proposed SVM-PoC chart and to compare it with other data mining-based control charts and Hotelling's T 2 control charts under various scenarios. The results showed that the proposed SVM-PoC charts outperformed other multivariate control charts in nonnormal situations. Further, we developed an exponential weighed moving average version of the SVM-PoC charts for increasing sensitivity to small shifts.

Original languageEnglish
Pages (from-to)1157-1173
Number of pages17
JournalJournal of Statistical Computation and Simulation
Volume81
Issue number9
DOIs
Publication statusPublished - 2011 Sep 1

Fingerprint

Process Monitoring
Process monitoring
Control Charts
Support vector machines
Support Vector Machine
Chart
Multivariate Control Charts
Monitoring
Process Capability
Statistical Process Control
Statistical process control
Bootstrap Method
Moving Average
Bootstrap
Statistic
Data mining
Control charts
Support vector machine
Diagnostics
Data Mining

Keywords

  • Bootstrap
  • Data mining
  • Multivariate control charts
  • Statistical quality control
  • Support vector machines

ASJC Scopus subject areas

  • Applied Mathematics
  • Statistics and Probability
  • Modelling and Simulation
  • Statistics, Probability and Uncertainty

Cite this

Integration of support vector machines and control charts for multivariate process monitoring. / Chongfuangprinya, Panitarn; Kim, Seoung Bum; Park, Sun Kyoung; Sukchotrat, Thuntee.

In: Journal of Statistical Computation and Simulation, Vol. 81, No. 9, 01.09.2011, p. 1157-1173.

Research output: Contribution to journalArticle

Chongfuangprinya, Panitarn ; Kim, Seoung Bum ; Park, Sun Kyoung ; Sukchotrat, Thuntee. / Integration of support vector machines and control charts for multivariate process monitoring. In: Journal of Statistical Computation and Simulation. 2011 ; Vol. 81, No. 9. pp. 1157-1173.
@article{2981d06944f44966aea44c50e307a4c1,
title = "Integration of support vector machines and control charts for multivariate process monitoring",
abstract = "Statistical process control tools have been used routinely to improve process capabilities through reliable on-line monitoring and diagnostic processes. In the present paper, we propose a novel multivariate control chart that integrates a support vector machine (SVM) algorithm, a bootstrap method, and a control chart technique to improve multivariate process monitoring. The proposed chart uses as the monitoring statistic the predicted probability of class (PoC) values from an SVM algorithm. The control limits of SVM-PoC charts are obtained by a bootstrap approach.A simulation study was conducted to evaluate the performance of the proposed SVM-PoC chart and to compare it with other data mining-based control charts and Hotelling's T 2 control charts under various scenarios. The results showed that the proposed SVM-PoC charts outperformed other multivariate control charts in nonnormal situations. Further, we developed an exponential weighed moving average version of the SVM-PoC charts for increasing sensitivity to small shifts.",
keywords = "Bootstrap, Data mining, Multivariate control charts, Statistical quality control, Support vector machines",
author = "Panitarn Chongfuangprinya and Kim, {Seoung Bum} and Park, {Sun Kyoung} and Thuntee Sukchotrat",
year = "2011",
month = "9",
day = "1",
doi = "10.1080/00949651003789074",
language = "English",
volume = "81",
pages = "1157--1173",
journal = "Journal of Statistical Computation and Simulation",
issn = "0094-9655",
publisher = "Taylor and Francis Ltd.",
number = "9",

}

TY - JOUR

T1 - Integration of support vector machines and control charts for multivariate process monitoring

AU - Chongfuangprinya, Panitarn

AU - Kim, Seoung Bum

AU - Park, Sun Kyoung

AU - Sukchotrat, Thuntee

PY - 2011/9/1

Y1 - 2011/9/1

N2 - Statistical process control tools have been used routinely to improve process capabilities through reliable on-line monitoring and diagnostic processes. In the present paper, we propose a novel multivariate control chart that integrates a support vector machine (SVM) algorithm, a bootstrap method, and a control chart technique to improve multivariate process monitoring. The proposed chart uses as the monitoring statistic the predicted probability of class (PoC) values from an SVM algorithm. The control limits of SVM-PoC charts are obtained by a bootstrap approach.A simulation study was conducted to evaluate the performance of the proposed SVM-PoC chart and to compare it with other data mining-based control charts and Hotelling's T 2 control charts under various scenarios. The results showed that the proposed SVM-PoC charts outperformed other multivariate control charts in nonnormal situations. Further, we developed an exponential weighed moving average version of the SVM-PoC charts for increasing sensitivity to small shifts.

AB - Statistical process control tools have been used routinely to improve process capabilities through reliable on-line monitoring and diagnostic processes. In the present paper, we propose a novel multivariate control chart that integrates a support vector machine (SVM) algorithm, a bootstrap method, and a control chart technique to improve multivariate process monitoring. The proposed chart uses as the monitoring statistic the predicted probability of class (PoC) values from an SVM algorithm. The control limits of SVM-PoC charts are obtained by a bootstrap approach.A simulation study was conducted to evaluate the performance of the proposed SVM-PoC chart and to compare it with other data mining-based control charts and Hotelling's T 2 control charts under various scenarios. The results showed that the proposed SVM-PoC charts outperformed other multivariate control charts in nonnormal situations. Further, we developed an exponential weighed moving average version of the SVM-PoC charts for increasing sensitivity to small shifts.

KW - Bootstrap

KW - Data mining

KW - Multivariate control charts

KW - Statistical quality control

KW - Support vector machines

UR - http://www.scopus.com/inward/record.url?scp=80051690841&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=80051690841&partnerID=8YFLogxK

U2 - 10.1080/00949651003789074

DO - 10.1080/00949651003789074

M3 - Article

AN - SCOPUS:80051690841

VL - 81

SP - 1157

EP - 1173

JO - Journal of Statistical Computation and Simulation

JF - Journal of Statistical Computation and Simulation

SN - 0094-9655

IS - 9

ER -