K-means clustering seeds initialization based on centrality, sparsity, and isotropy

Pilsung Kang, Sungzoon Cho

Research output: Chapter in Book/Report/Conference proceedingConference contribution

9 Citations (Scopus)

Abstract

K-Means is the most commonly used clustering algorithm. Despite its numerous advantages, it has a crucial drawback: the final cluster structure entirely relies on the choice of initial seeds. In this paper, a new seeds initialization algorithm based on centrality, sparsity, and isotropy is proposed. Preliminary experiments show that the proposed algorithm not only resulted in better clustering structures, but also accelerated the convergence.

Original languageEnglish
Title of host publicationIntelligent Data Engineering and Automated Learning - IDEAL 2009 - 10th International Conference, Proceedings
Pages109-117
Number of pages9
DOIs
Publication statusPublished - 2009 Dec 1
Externally publishedYes
Event10th International Conference on Intelligent Data Engineering and Automated Learning, IDEAL 2009 - Burgos, Spain
Duration: 2009 Sep 232009 Sep 26

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume5788 LNCS
ISSN (Print)0302-9743
ISSN (Electronic)1611-3349

Conference

Conference10th International Conference on Intelligent Data Engineering and Automated Learning, IDEAL 2009
CountrySpain
CityBurgos
Period09/9/2309/9/26

Fingerprint

Centrality
K-means Clustering
Isotropy
Initialization
Sparsity
Seed
K-means
Clustering algorithms
Clustering Algorithm
Clustering
Experiment
Experiments

ASJC Scopus subject areas

  • Computer Science(all)
  • Theoretical Computer Science

Cite this

Kang, P., & Cho, S. (2009). K-means clustering seeds initialization based on centrality, sparsity, and isotropy. In Intelligent Data Engineering and Automated Learning - IDEAL 2009 - 10th International Conference, Proceedings (pp. 109-117). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 5788 LNCS). https://doi.org/10.1007/978-3-642-04394-9_14

K-means clustering seeds initialization based on centrality, sparsity, and isotropy. / Kang, Pilsung; Cho, Sungzoon.

Intelligent Data Engineering and Automated Learning - IDEAL 2009 - 10th International Conference, Proceedings. 2009. p. 109-117 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 5788 LNCS).

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Kang, P & Cho, S 2009, K-means clustering seeds initialization based on centrality, sparsity, and isotropy. in Intelligent Data Engineering and Automated Learning - IDEAL 2009 - 10th International Conference, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 5788 LNCS, pp. 109-117, 10th International Conference on Intelligent Data Engineering and Automated Learning, IDEAL 2009, Burgos, Spain, 09/9/23. https://doi.org/10.1007/978-3-642-04394-9_14
Kang P, Cho S. K-means clustering seeds initialization based on centrality, sparsity, and isotropy. In Intelligent Data Engineering and Automated Learning - IDEAL 2009 - 10th International Conference, Proceedings. 2009. p. 109-117. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). https://doi.org/10.1007/978-3-642-04394-9_14
Kang, Pilsung ; Cho, Sungzoon. / K-means clustering seeds initialization based on centrality, sparsity, and isotropy. Intelligent Data Engineering and Automated Learning - IDEAL 2009 - 10th International Conference, Proceedings. 2009. pp. 109-117 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{acca4ef5994e47b9a47c8a3373bdd980,
title = "K-means clustering seeds initialization based on centrality, sparsity, and isotropy",
abstract = "K-Means is the most commonly used clustering algorithm. Despite its numerous advantages, it has a crucial drawback: the final cluster structure entirely relies on the choice of initial seeds. In this paper, a new seeds initialization algorithm based on centrality, sparsity, and isotropy is proposed. Preliminary experiments show that the proposed algorithm not only resulted in better clustering structures, but also accelerated the convergence.",
author = "Pilsung Kang and Sungzoon Cho",
year = "2009",
month = "12",
day = "1",
doi = "10.1007/978-3-642-04394-9_14",
language = "English",
isbn = "3642043933",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
pages = "109--117",
booktitle = "Intelligent Data Engineering and Automated Learning - IDEAL 2009 - 10th International Conference, Proceedings",

}

TY - GEN

T1 - K-means clustering seeds initialization based on centrality, sparsity, and isotropy

AU - Kang, Pilsung

AU - Cho, Sungzoon

PY - 2009/12/1

Y1 - 2009/12/1

N2 - K-Means is the most commonly used clustering algorithm. Despite its numerous advantages, it has a crucial drawback: the final cluster structure entirely relies on the choice of initial seeds. In this paper, a new seeds initialization algorithm based on centrality, sparsity, and isotropy is proposed. Preliminary experiments show that the proposed algorithm not only resulted in better clustering structures, but also accelerated the convergence.

AB - K-Means is the most commonly used clustering algorithm. Despite its numerous advantages, it has a crucial drawback: the final cluster structure entirely relies on the choice of initial seeds. In this paper, a new seeds initialization algorithm based on centrality, sparsity, and isotropy is proposed. Preliminary experiments show that the proposed algorithm not only resulted in better clustering structures, but also accelerated the convergence.

UR - http://www.scopus.com/inward/record.url?scp=76249127819&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=76249127819&partnerID=8YFLogxK

U2 - 10.1007/978-3-642-04394-9_14

DO - 10.1007/978-3-642-04394-9_14

M3 - Conference contribution

SN - 3642043933

SN - 9783642043932

T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)

SP - 109

EP - 117

BT - Intelligent Data Engineering and Automated Learning - IDEAL 2009 - 10th International Conference, Proceedings

ER -