Ordered multiple-class ROC analysis with continuous measurements

Christos T. Nakas, Constantin Yiannoutsos

Research output: Contribution to journalArticle

133 Citations (Scopus)

Abstract

Receiver operating characteristic (ROC) curves have been useful in two-group classification problems. In three- and multiple-class diagnostic problems, an ROC surface or hyper-surface can be constructed. The volume under these surfaces can be used for inference using bootstrap techniques or U-statistics theory. In this article, ROC surfaces and hyper-surfaces are defined and their behaviour and utility in multi-group classification problems is investigated. The formulation of the problem is equivalent to what has previously been proposed in the general multi-category classification problem but the definition of ROC surfaces here is less complex and addresses directly the narrower problem of ordered categories in the three-class and, by extension, the multi-class problem applied to continuous and ordinal data. Non-parametric manipulation of both continuous and discrete test data and comparison between two diagnostic tests applied to the same subjects are considered. A three-group classification example in the context of HIV neurological disease is presented and the results are discussed.

Original languageEnglish
Pages (from-to)3437-3449
Number of pages13
JournalStatistics in Medicine
Volume23
Issue number22
DOIs
StatePublished - Nov 30 2004

Fingerprint

Operating Characteristics
Group Classification
ROC Curve
Receiver
Classification Problems
Hypersurface
Ordered Categories
Ordinal Data
Diagnostic Tests
U-statistics
Receiver Operating Characteristic Curve
Multi-class
Routine Diagnostic Tests
Bootstrap
Manipulation
Diagnostics
Research Design
HIV
Class
Formulation

Keywords

  • Bootstrap
  • Diagnostic testing
  • Neuropsychological testing
  • ROC analysis
  • U-statistics

ASJC Scopus subject areas

  • Epidemiology

Cite this

Ordered multiple-class ROC analysis with continuous measurements. / Nakas, Christos T.; Yiannoutsos, Constantin.

In: Statistics in Medicine, Vol. 23, No. 22, 30.11.2004, p. 3437-3449.

Research output: Contribution to journalArticle

@article{bccf0f4c431647079163c10470b2bd9b,
title = "Ordered multiple-class ROC analysis with continuous measurements",
abstract = "Receiver operating characteristic (ROC) curves have been useful in two-group classification problems. In three- and multiple-class diagnostic problems, an ROC surface or hyper-surface can be constructed. The volume under these surfaces can be used for inference using bootstrap techniques or U-statistics theory. In this article, ROC surfaces and hyper-surfaces are defined and their behaviour and utility in multi-group classification problems is investigated. The formulation of the problem is equivalent to what has previously been proposed in the general multi-category classification problem but the definition of ROC surfaces here is less complex and addresses directly the narrower problem of ordered categories in the three-class and, by extension, the multi-class problem applied to continuous and ordinal data. Non-parametric manipulation of both continuous and discrete test data and comparison between two diagnostic tests applied to the same subjects are considered. A three-group classification example in the context of HIV neurological disease is presented and the results are discussed.",
keywords = "Bootstrap, Diagnostic testing, Neuropsychological testing, ROC analysis, U-statistics",
author = "Nakas, {Christos T.} and Constantin Yiannoutsos",
year = "2004",
month = "11",
day = "30",
doi = "10.1002/sim.1917",
language = "English",
volume = "23",
pages = "3437--3449",
journal = "Statistics in Medicine",
issn = "0277-6715",
publisher = "John Wiley and Sons Ltd",
number = "22",

}

TY - JOUR

T1 - Ordered multiple-class ROC analysis with continuous measurements

AU - Nakas, Christos T.

AU - Yiannoutsos, Constantin

PY - 2004/11/30

Y1 - 2004/11/30

N2 - Receiver operating characteristic (ROC) curves have been useful in two-group classification problems. In three- and multiple-class diagnostic problems, an ROC surface or hyper-surface can be constructed. The volume under these surfaces can be used for inference using bootstrap techniques or U-statistics theory. In this article, ROC surfaces and hyper-surfaces are defined and their behaviour and utility in multi-group classification problems is investigated. The formulation of the problem is equivalent to what has previously been proposed in the general multi-category classification problem but the definition of ROC surfaces here is less complex and addresses directly the narrower problem of ordered categories in the three-class and, by extension, the multi-class problem applied to continuous and ordinal data. Non-parametric manipulation of both continuous and discrete test data and comparison between two diagnostic tests applied to the same subjects are considered. A three-group classification example in the context of HIV neurological disease is presented and the results are discussed.

AB - Receiver operating characteristic (ROC) curves have been useful in two-group classification problems. In three- and multiple-class diagnostic problems, an ROC surface or hyper-surface can be constructed. The volume under these surfaces can be used for inference using bootstrap techniques or U-statistics theory. In this article, ROC surfaces and hyper-surfaces are defined and their behaviour and utility in multi-group classification problems is investigated. The formulation of the problem is equivalent to what has previously been proposed in the general multi-category classification problem but the definition of ROC surfaces here is less complex and addresses directly the narrower problem of ordered categories in the three-class and, by extension, the multi-class problem applied to continuous and ordinal data. Non-parametric manipulation of both continuous and discrete test data and comparison between two diagnostic tests applied to the same subjects are considered. A three-group classification example in the context of HIV neurological disease is presented and the results are discussed.

KW - Bootstrap

KW - Diagnostic testing

KW - Neuropsychological testing

KW - ROC analysis

KW - U-statistics

UR - http://www.scopus.com/inward/record.url?scp=7944220109&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=7944220109&partnerID=8YFLogxK

U2 - 10.1002/sim.1917

DO - 10.1002/sim.1917

M3 - Article

C2 - 15505886

AN - SCOPUS:7944220109

VL - 23

SP - 3437

EP - 3449

JO - Statistics in Medicine

JF - Statistics in Medicine

SN - 0277-6715

IS - 22

ER -