Comparison of novelty score-based multivariate control charts

Gulanbaier Tuerhong, Seoung Bum Kim

Research output: Contribution to journalArticle

6 Citations (Scopus)

Abstract

Control charts are widely used in various industries to improve product quality. One recent trend in developing control charts is based on novelty score algorithms that can effectively describe reality and reflect the unique characteristics of the data being monitored. In this study, we compared eight novelty score algorithms - the T2, Local T2, Dmax, Dmean, K2, the k-nearest neighbor data description, the local density outlier factor, and the hybrid novelty score (HNS) - in terms of their average run length performance. A rigorous simulation was conducted to compare the novelty score-based multivariate control charts under both normal and non-normal scenarios. The simulation showed that in both normal and lognormal scenarios, Dmax-based control charts produced the most promising results. In skewed distribution with high kurtosis non-normal scenarios, HNS- and K2-based control charts performed best. In symmetric with kurtosis non-normal scenarios, local T2-based control charts outperformed the others.

Original languageEnglish
Pages (from-to)1126-1143
Number of pages18
JournalCommunications in Statistics: Simulation and Computation
Volume44
Issue number5
DOIs
Publication statusPublished - 2015 Jan 1

Fingerprint

Multivariate Control Charts
Control Charts
Scenarios
Kurtosis
Average Run Length
Skewed Distribution
Data description
Outlier
Nearest Neighbor
Simulation
Control charts
Industry

ASJC Scopus subject areas

  • Modelling and Simulation
  • Statistics and Probability

Cite this

Comparison of novelty score-based multivariate control charts. / Tuerhong, Gulanbaier; Kim, Seoung Bum.

In: Communications in Statistics: Simulation and Computation, Vol. 44, No. 5, 01.01.2015, p. 1126-1143.

Research output: Contribution to journalArticle

@article{f112782497e1405c9deb14a448cf33ee,
title = "Comparison of novelty score-based multivariate control charts",
abstract = "Control charts are widely used in various industries to improve product quality. One recent trend in developing control charts is based on novelty score algorithms that can effectively describe reality and reflect the unique characteristics of the data being monitored. In this study, we compared eight novelty score algorithms - the T2, Local T2, Dmax, Dmean, K2, the k-nearest neighbor data description, the local density outlier factor, and the hybrid novelty score (HNS) - in terms of their average run length performance. A rigorous simulation was conducted to compare the novelty score-based multivariate control charts under both normal and non-normal scenarios. The simulation showed that in both normal and lognormal scenarios, Dmax-based control charts produced the most promising results. In skewed distribution with high kurtosis non-normal scenarios, HNS- and K2-based control charts performed best. In symmetric with kurtosis non-normal scenarios, local T2-based control charts outperformed the others.",
keywords = "Data mining, Multivariate control charts, Novelty score, Quality control.",
author = "Gulanbaier Tuerhong and Kim, {Seoung Bum}",
year = "2015",
month = "1",
day = "1",
doi = "10.1080/03610918.2013.809098",
language = "English",
volume = "44",
pages = "1126--1143",
journal = "Communications in Statistics Part B: Simulation and Computation",
issn = "0361-0918",
publisher = "Taylor and Francis Ltd.",
number = "5",

}

TY - JOUR

T1 - Comparison of novelty score-based multivariate control charts

AU - Tuerhong, Gulanbaier

AU - Kim, Seoung Bum

PY - 2015/1/1

Y1 - 2015/1/1

N2 - Control charts are widely used in various industries to improve product quality. One recent trend in developing control charts is based on novelty score algorithms that can effectively describe reality and reflect the unique characteristics of the data being monitored. In this study, we compared eight novelty score algorithms - the T2, Local T2, Dmax, Dmean, K2, the k-nearest neighbor data description, the local density outlier factor, and the hybrid novelty score (HNS) - in terms of their average run length performance. A rigorous simulation was conducted to compare the novelty score-based multivariate control charts under both normal and non-normal scenarios. The simulation showed that in both normal and lognormal scenarios, Dmax-based control charts produced the most promising results. In skewed distribution with high kurtosis non-normal scenarios, HNS- and K2-based control charts performed best. In symmetric with kurtosis non-normal scenarios, local T2-based control charts outperformed the others.

AB - Control charts are widely used in various industries to improve product quality. One recent trend in developing control charts is based on novelty score algorithms that can effectively describe reality and reflect the unique characteristics of the data being monitored. In this study, we compared eight novelty score algorithms - the T2, Local T2, Dmax, Dmean, K2, the k-nearest neighbor data description, the local density outlier factor, and the hybrid novelty score (HNS) - in terms of their average run length performance. A rigorous simulation was conducted to compare the novelty score-based multivariate control charts under both normal and non-normal scenarios. The simulation showed that in both normal and lognormal scenarios, Dmax-based control charts produced the most promising results. In skewed distribution with high kurtosis non-normal scenarios, HNS- and K2-based control charts performed best. In symmetric with kurtosis non-normal scenarios, local T2-based control charts outperformed the others.

KW - Data mining

KW - Multivariate control charts

KW - Novelty score

KW - Quality control.

UR - http://www.scopus.com/inward/record.url?scp=84908621006&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84908621006&partnerID=8YFLogxK

U2 - 10.1080/03610918.2013.809098

DO - 10.1080/03610918.2013.809098

M3 - Article

VL - 44

SP - 1126

EP - 1143

JO - Communications in Statistics Part B: Simulation and Computation

JF - Communications in Statistics Part B: Simulation and Computation

SN - 0361-0918

IS - 5

ER -