An extended projection data depth and its applications to discrimination

Xia Cui, Lu Lin, Guangren Yang

Research output: Contribution to journalArticle

14 Citations (Scopus)

Abstract

This article investigates the possible use of our newly defined extended projection depth (abbreviated to EPD) in nonparametric discriminant analysis. We propose a robust nonparametric classifier, which relies on the intuitively simple notion of EPD. The EPD-based classifier assigns an observation to the population with respect to which it has the maximum EPD. Asymptotic properties of misclassification rates and robust properties of EPD-based classifier are discussed. A few simulated data sets are used to compare the performance of EPD-based classifier with Fisher's linear discriminant rule, quadratic discriminant rule, and PD-based classifier. It is also found that when the underlying distributions are elliptically symmetric, EPD-based classifier is asymptotically equivalent to the optimal Bayes classifier.

Original languageEnglish (US)
Pages (from-to)2276-2290
Number of pages15
JournalCommunications in Statistics - Theory and Methods
Volume37
Issue number14
DOIs
StatePublished - Jan 1 2008

Fingerprint

Data Depth
Discrimination
Classifier
Projection
Discriminant
Bayes Classifier
Misclassification Rate
Asymptotically equivalent
Discriminant Analysis
Asymptotic Properties
Assign

All Science Journal Classification (ASJC) codes

  • Statistics and Probability

Cite this

Cui, Xia ; Lin, Lu ; Yang, Guangren. / An extended projection data depth and its applications to discrimination. In: Communications in Statistics - Theory and Methods. 2008 ; Vol. 37, No. 14. pp. 2276-2290.
@article{0dbf160c1ce1434f9d1a406ba4322880,
title = "An extended projection data depth and its applications to discrimination",
abstract = "This article investigates the possible use of our newly defined extended projection depth (abbreviated to EPD) in nonparametric discriminant analysis. We propose a robust nonparametric classifier, which relies on the intuitively simple notion of EPD. The EPD-based classifier assigns an observation to the population with respect to which it has the maximum EPD. Asymptotic properties of misclassification rates and robust properties of EPD-based classifier are discussed. A few simulated data sets are used to compare the performance of EPD-based classifier with Fisher's linear discriminant rule, quadratic discriminant rule, and PD-based classifier. It is also found that when the underlying distributions are elliptically symmetric, EPD-based classifier is asymptotically equivalent to the optimal Bayes classifier.",
author = "Xia Cui and Lu Lin and Guangren Yang",
year = "2008",
month = "1",
day = "1",
doi = "10.1080/03610920701858396",
language = "English (US)",
volume = "37",
pages = "2276--2290",
journal = "Communications in Statistics - Theory and Methods",
issn = "0361-0926",
publisher = "Taylor and Francis Ltd.",
number = "14",

}

An extended projection data depth and its applications to discrimination. / Cui, Xia; Lin, Lu; Yang, Guangren.

In: Communications in Statistics - Theory and Methods, Vol. 37, No. 14, 01.01.2008, p. 2276-2290.

Research output: Contribution to journalArticle

TY - JOUR

T1 - An extended projection data depth and its applications to discrimination

AU - Cui, Xia

AU - Lin, Lu

AU - Yang, Guangren

PY - 2008/1/1

Y1 - 2008/1/1

N2 - This article investigates the possible use of our newly defined extended projection depth (abbreviated to EPD) in nonparametric discriminant analysis. We propose a robust nonparametric classifier, which relies on the intuitively simple notion of EPD. The EPD-based classifier assigns an observation to the population with respect to which it has the maximum EPD. Asymptotic properties of misclassification rates and robust properties of EPD-based classifier are discussed. A few simulated data sets are used to compare the performance of EPD-based classifier with Fisher's linear discriminant rule, quadratic discriminant rule, and PD-based classifier. It is also found that when the underlying distributions are elliptically symmetric, EPD-based classifier is asymptotically equivalent to the optimal Bayes classifier.

AB - This article investigates the possible use of our newly defined extended projection depth (abbreviated to EPD) in nonparametric discriminant analysis. We propose a robust nonparametric classifier, which relies on the intuitively simple notion of EPD. The EPD-based classifier assigns an observation to the population with respect to which it has the maximum EPD. Asymptotic properties of misclassification rates and robust properties of EPD-based classifier are discussed. A few simulated data sets are used to compare the performance of EPD-based classifier with Fisher's linear discriminant rule, quadratic discriminant rule, and PD-based classifier. It is also found that when the underlying distributions are elliptically symmetric, EPD-based classifier is asymptotically equivalent to the optimal Bayes classifier.

UR - http://www.scopus.com/inward/record.url?scp=46249108557&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=46249108557&partnerID=8YFLogxK

U2 - 10.1080/03610920701858396

DO - 10.1080/03610920701858396

M3 - Article

AN - SCOPUS:46249108557

VL - 37

SP - 2276

EP - 2290

JO - Communications in Statistics - Theory and Methods

JF - Communications in Statistics - Theory and Methods

SN - 0361-0926

IS - 14

ER -