Hard and fuzzy c-means clustering with conditionally positive definite kernel

Yuchi Kanzawa, Yasunori Yasunori Endo, Sadaaki Miyamoto

Research output: Contribution to journalArticle

Abstract

In this paper, we investigate three types of c-means clustering algorithms with a conditionally positive definite (cpd) kernel. One is based on hard c-means and two are based on standard and entropy-regularized fuzzy c-means. First, based on a cpd kernel describing a squared Euclidean distance between data in feature space, these algorithms are derived from revised optimization problems of the conventional kernel c-means. Next, based on the relationship between the positive definite (pd) kernel and cpd kernel, the revised dissimilarity between a datum and a cluster center in the feature space is shown. Finally, it is shown that a cpd kernel c-means algorithm and a kernel c-means algorithm with a pd kernel derived from the cpd kernel are essentially identical to each other. Explicit mapping for a cpd kernel is also described geometrically.

Original languageEnglish
Pages (from-to)825-830
Number of pages6
JournalJournal of Advanced Computational Intelligence and Intelligent Informatics
Volume16
Issue number7
Publication statusPublished - 2012 Nov

Fingerprint

Clustering algorithms
Entropy

Keywords

  • Clustering
  • Conditionally positive definite kernel
  • Fuzzy c-means

ASJC Scopus subject areas

  • Artificial Intelligence
  • Computer Vision and Pattern Recognition
  • Human-Computer Interaction

Cite this

Hard and fuzzy c-means clustering with conditionally positive definite kernel. / Kanzawa, Yuchi; Yasunori Endo, Yasunori; Miyamoto, Sadaaki.

In: Journal of Advanced Computational Intelligence and Intelligent Informatics, Vol. 16, No. 7, 11.2012, p. 825-830.

Research output: Contribution to journalArticle

@article{dcec32280c1d406e8327b83c613aaf75,
title = "Hard and fuzzy c-means clustering with conditionally positive definite kernel",
abstract = "In this paper, we investigate three types of c-means clustering algorithms with a conditionally positive definite (cpd) kernel. One is based on hard c-means and two are based on standard and entropy-regularized fuzzy c-means. First, based on a cpd kernel describing a squared Euclidean distance between data in feature space, these algorithms are derived from revised optimization problems of the conventional kernel c-means. Next, based on the relationship between the positive definite (pd) kernel and cpd kernel, the revised dissimilarity between a datum and a cluster center in the feature space is shown. Finally, it is shown that a cpd kernel c-means algorithm and a kernel c-means algorithm with a pd kernel derived from the cpd kernel are essentially identical to each other. Explicit mapping for a cpd kernel is also described geometrically.",
keywords = "Clustering, Conditionally positive definite kernel, Fuzzy c-means",
author = "Yuchi Kanzawa and {Yasunori Endo}, Yasunori and Sadaaki Miyamoto",
year = "2012",
month = "11",
language = "English",
volume = "16",
pages = "825--830",
journal = "Journal of Advanced Computational Intelligence and Intelligent Informatics",
issn = "1343-0130",
publisher = "Fuji Technology Press",
number = "7",

}

TY - JOUR

T1 - Hard and fuzzy c-means clustering with conditionally positive definite kernel

AU - Kanzawa, Yuchi

AU - Yasunori Endo, Yasunori

AU - Miyamoto, Sadaaki

PY - 2012/11

Y1 - 2012/11

N2 - In this paper, we investigate three types of c-means clustering algorithms with a conditionally positive definite (cpd) kernel. One is based on hard c-means and two are based on standard and entropy-regularized fuzzy c-means. First, based on a cpd kernel describing a squared Euclidean distance between data in feature space, these algorithms are derived from revised optimization problems of the conventional kernel c-means. Next, based on the relationship between the positive definite (pd) kernel and cpd kernel, the revised dissimilarity between a datum and a cluster center in the feature space is shown. Finally, it is shown that a cpd kernel c-means algorithm and a kernel c-means algorithm with a pd kernel derived from the cpd kernel are essentially identical to each other. Explicit mapping for a cpd kernel is also described geometrically.

AB - In this paper, we investigate three types of c-means clustering algorithms with a conditionally positive definite (cpd) kernel. One is based on hard c-means and two are based on standard and entropy-regularized fuzzy c-means. First, based on a cpd kernel describing a squared Euclidean distance between data in feature space, these algorithms are derived from revised optimization problems of the conventional kernel c-means. Next, based on the relationship between the positive definite (pd) kernel and cpd kernel, the revised dissimilarity between a datum and a cluster center in the feature space is shown. Finally, it is shown that a cpd kernel c-means algorithm and a kernel c-means algorithm with a pd kernel derived from the cpd kernel are essentially identical to each other. Explicit mapping for a cpd kernel is also described geometrically.

KW - Clustering

KW - Conditionally positive definite kernel

KW - Fuzzy c-means

UR - http://www.scopus.com/inward/record.url?scp=84872460178&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84872460178&partnerID=8YFLogxK

M3 - Article

AN - SCOPUS:84872460178

VL - 16

SP - 825

EP - 830

JO - Journal of Advanced Computational Intelligence and Intelligent Informatics

JF - Journal of Advanced Computational Intelligence and Intelligent Informatics

SN - 1343-0130

IS - 7

ER -