Test-based classification: A linkage between classification and statistical testing

Shu Min Liao, Michael Akritas

Research output: Contribution to journalArticle

6 Citations (Scopus)

Abstract

The purpose of this article is to introduce a new classification methodology. The methodology uses a connection, which we uncover, between classification and testing, and is called Test-based classification. Although the main focus of this article is the binary classification with the univariate and the multivariate data, an extension to the multiclass classification is also covered. Several simulated and real data sets are used to demonstrate how this new methodology works. We argue that our new idea is competitive with the linear and quadratic discriminant analysis when the observed data are normally distributed, but it can outperform them when the data are not normally distributed. Lanchenbruch's holdout misclassification rate is used to judge the performance of classification.

Original languageEnglish (US)
Pages (from-to)1269-1281
Number of pages13
JournalStatistics and Probability Letters
Volume77
Issue number12
DOIs
StatePublished - Jul 1 2007

Fingerprint

Linkage
Testing
Methodology
Misclassification Rate
Multi-class Classification
Binary Classification
Multivariate Data
Discriminant Analysis
Univariate
Statistical testing
Demonstrate

All Science Journal Classification (ASJC) codes

  • Statistics and Probability
  • Statistics, Probability and Uncertainty

Cite this

@article{f442ca63d99849b298ae073c37e8746d,
title = "Test-based classification: A linkage between classification and statistical testing",
abstract = "The purpose of this article is to introduce a new classification methodology. The methodology uses a connection, which we uncover, between classification and testing, and is called Test-based classification. Although the main focus of this article is the binary classification with the univariate and the multivariate data, an extension to the multiclass classification is also covered. Several simulated and real data sets are used to demonstrate how this new methodology works. We argue that our new idea is competitive with the linear and quadratic discriminant analysis when the observed data are normally distributed, but it can outperform them when the data are not normally distributed. Lanchenbruch's holdout misclassification rate is used to judge the performance of classification.",
author = "Liao, {Shu Min} and Michael Akritas",
year = "2007",
month = "7",
day = "1",
doi = "10.1016/j.spl.2007.03.012",
language = "English (US)",
volume = "77",
pages = "1269--1281",
journal = "Statistics and Probability Letters",
issn = "0167-7152",
publisher = "Elsevier",
number = "12",

}

Test-based classification : A linkage between classification and statistical testing. / Liao, Shu Min; Akritas, Michael.

In: Statistics and Probability Letters, Vol. 77, No. 12, 01.07.2007, p. 1269-1281.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Test-based classification

T2 - A linkage between classification and statistical testing

AU - Liao, Shu Min

AU - Akritas, Michael

PY - 2007/7/1

Y1 - 2007/7/1

N2 - The purpose of this article is to introduce a new classification methodology. The methodology uses a connection, which we uncover, between classification and testing, and is called Test-based classification. Although the main focus of this article is the binary classification with the univariate and the multivariate data, an extension to the multiclass classification is also covered. Several simulated and real data sets are used to demonstrate how this new methodology works. We argue that our new idea is competitive with the linear and quadratic discriminant analysis when the observed data are normally distributed, but it can outperform them when the data are not normally distributed. Lanchenbruch's holdout misclassification rate is used to judge the performance of classification.

AB - The purpose of this article is to introduce a new classification methodology. The methodology uses a connection, which we uncover, between classification and testing, and is called Test-based classification. Although the main focus of this article is the binary classification with the univariate and the multivariate data, an extension to the multiclass classification is also covered. Several simulated and real data sets are used to demonstrate how this new methodology works. We argue that our new idea is competitive with the linear and quadratic discriminant analysis when the observed data are normally distributed, but it can outperform them when the data are not normally distributed. Lanchenbruch's holdout misclassification rate is used to judge the performance of classification.

UR - http://www.scopus.com/inward/record.url?scp=34249895313&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=34249895313&partnerID=8YFLogxK

U2 - 10.1016/j.spl.2007.03.012

DO - 10.1016/j.spl.2007.03.012

M3 - Article

AN - SCOPUS:34249895313

VL - 77

SP - 1269

EP - 1281

JO - Statistics and Probability Letters

JF - Statistics and Probability Letters

SN - 0167-7152

IS - 12

ER -