Determining the number of clusters in cluster analysis

My Young Cheong, Hakbae Lee

Research output: Contribution to journalArticle

11 Citations (Scopus)

Abstract

Cluster analysis has been a popular method for statistical classification. The classical cluster analysis, however, has a theoretical shortcoming in the sense that the inference to determine the number of clusters does not provide the theoretical guideline. To estimate the number of clusters, this paper explores the problem through the EM algorithm, Maximum a Posteriori and Gibbs sampler. In addition, we investigate the Bayesian Information criteria (BIC), the Laplace Metropolis criteria and the modified Fisher's criteria in order to determine the number of clusters.

Original languageEnglish
Pages (from-to)135-143
Number of pages9
JournalJournal of the Korean Statistical Society
Volume37
Issue number2
DOIs
Publication statusPublished - 2008 Jun 1

Fingerprint

Number of Clusters
Cluster Analysis
Bayesian Information Criterion
Gibbs Sampler
Maximum a Posteriori
EM Algorithm
Laplace
Estimate

All Science Journal Classification (ASJC) codes

  • Statistics and Probability

Cite this

@article{2622682eddd44aaba760a5bc399a28ec,
title = "Determining the number of clusters in cluster analysis",
abstract = "Cluster analysis has been a popular method for statistical classification. The classical cluster analysis, however, has a theoretical shortcoming in the sense that the inference to determine the number of clusters does not provide the theoretical guideline. To estimate the number of clusters, this paper explores the problem through the EM algorithm, Maximum a Posteriori and Gibbs sampler. In addition, we investigate the Bayesian Information criteria (BIC), the Laplace Metropolis criteria and the modified Fisher's criteria in order to determine the number of clusters.",
author = "Cheong, {My Young} and Hakbae Lee",
year = "2008",
month = "6",
day = "1",
doi = "10.1016/j.jkss.2007.10.004",
language = "English",
volume = "37",
pages = "135--143",
journal = "Journal of the Korean Statistical Society",
issn = "1226-3192",
publisher = "Korean Statistical Society",
number = "2",

}

Determining the number of clusters in cluster analysis. / Cheong, My Young; Lee, Hakbae.

In: Journal of the Korean Statistical Society, Vol. 37, No. 2, 01.06.2008, p. 135-143.

Research output: Contribution to journalArticle

TY - JOUR

T1 - Determining the number of clusters in cluster analysis

AU - Cheong, My Young

AU - Lee, Hakbae

PY - 2008/6/1

Y1 - 2008/6/1

N2 - Cluster analysis has been a popular method for statistical classification. The classical cluster analysis, however, has a theoretical shortcoming in the sense that the inference to determine the number of clusters does not provide the theoretical guideline. To estimate the number of clusters, this paper explores the problem through the EM algorithm, Maximum a Posteriori and Gibbs sampler. In addition, we investigate the Bayesian Information criteria (BIC), the Laplace Metropolis criteria and the modified Fisher's criteria in order to determine the number of clusters.

AB - Cluster analysis has been a popular method for statistical classification. The classical cluster analysis, however, has a theoretical shortcoming in the sense that the inference to determine the number of clusters does not provide the theoretical guideline. To estimate the number of clusters, this paper explores the problem through the EM algorithm, Maximum a Posteriori and Gibbs sampler. In addition, we investigate the Bayesian Information criteria (BIC), the Laplace Metropolis criteria and the modified Fisher's criteria in order to determine the number of clusters.

UR - http://www.scopus.com/inward/record.url?scp=43049127228&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=43049127228&partnerID=8YFLogxK

U2 - 10.1016/j.jkss.2007.10.004

DO - 10.1016/j.jkss.2007.10.004

M3 - Article

VL - 37

SP - 135

EP - 143

JO - Journal of the Korean Statistical Society

JF - Journal of the Korean Statistical Society

SN - 1226-3192

IS - 2

ER -