High-dimensional Data Clustering with Fuzzy C-Means: Problem, Reason, and Solution

Yinghua Shen; Hanyu E; Tianhua Chen; Zhi Xiao; Bingsheng Liu; Yuan Chen

doi:10.1007/978-3-030-85030-2_8

High-dimensional Data Clustering with Fuzzy C-Means: Problem, Reason, and Solution

Yinghua Shen, Hanyu E, Tianhua Chen, Zhi Xiao, Bingsheng Liu, Yuan Chen

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

3 Citations (Scopus)

Abstract

Fuzzy C-Means (FCM) clustering algorithm is a popular unsupervised learning approach that has been extensively utilized in various domains. However, in this study, we point out a major problem faced by FCM when it is applied to the high-dimensional data, i.e., quite often the obtained prototypes (cluster centers) could not be distinguished with each other. Many studies have claimed that the concentration of the distance (CoD) could be a major reason for this phenomenon. This paper has therefore revisited this factor, and highlight that the CoD could not only lead to decreased performance, but sometimes also positively contribute to enhanced performance of the clustering algorithm. Instead, this paper point out the significance of features that are noisy and correlated, which could have a negative effect on FCM performance. Hence, to tackle the mentioned problem, we resort to a neural network model, i.e., the autoencoder, to reduce the dimensionality of the feature space while extracting features that are most informative. We conduct several experiments to show the validity of the proposed strategy for the FCM algorithm.

Original language	English
Title of host publication	Advances in Computational Intelligence
Subtitle of host publication	16th International Work-Conference on Artificial Neural Networks, IWANN 2021, Virtual Event, June 16–18, 2021, Proceedings, Part I
Editors	Ignacio Rojas, Gonzalo Joya, Andreu Català
Place of Publication	Cham
Publisher	Springer Nature Switzerland AG
Pages	89-100
Number of pages	12
Volume	LNCS/LNTCS 12861
Edition	1st
ISBN (Electronic)	9783030850302
ISBN (Print)	9783030850296
DOIs	https://doi.org/10.1007/978-3-030-85030-2_8
Publication status	Published - 1 Sep 2021
Event	16th International Work-Conference on Artificial Neural Networks - Virtual Duration: 16 Jun 2021 → 18 Jun 2021 Conference number: 16 http://iwann.uma.es/

Publication series

Name	Lecture Notes in Computer Science
Publisher	Springer Nature Switzerland AG
Volume	LNCS/LNTCS 12861
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	16th International Work-Conference on Artificial Neural Networks
Abbreviated title	IWANN 2021
City	Virtual
Period	16/06/21 → 18/06/21
Internet address	http://iwann.uma.es/

Access to Document

10.1007/978-3-030-85030-2_8Licence: Unspecified

paper_17Accepted author manuscript, 1.55 MBLicence: Unspecified

Cite this

Shen, Y., E, H., Chen, T., Xiao, Z., Liu, B., & Chen, Y. (2021). High-dimensional Data Clustering with Fuzzy C-Means: Problem, Reason, and Solution. In I. Rojas, G. Joya, & A. Català (Eds.), Advances in Computational Intelligence: 16th International Work-Conference on Artificial Neural Networks, IWANN 2021, Virtual Event, June 16–18, 2021, Proceedings, Part I (1st ed., Vol. LNCS/LNTCS 12861, pp. 89-100). (Lecture Notes in Computer Science; Vol. LNCS/LNTCS 12861). Springer Nature Switzerland AG. https://doi.org/10.1007/978-3-030-85030-2_8

Shen, Yinghua ; E, Hanyu ; Chen, Tianhua et al. / High-dimensional Data Clustering with Fuzzy C-Means : Problem, Reason, and Solution. Advances in Computational Intelligence: 16th International Work-Conference on Artificial Neural Networks, IWANN 2021, Virtual Event, June 16–18, 2021, Proceedings, Part I. editor / Ignacio Rojas ; Gonzalo Joya ; Andreu Català. Vol. LNCS/LNTCS 12861 1st. ed. Cham : Springer Nature Switzerland AG, 2021. pp. 89-100 (Lecture Notes in Computer Science).

@inproceedings{4b0d2bbb9ef841ea91be18185bee8145,

title = "High-dimensional Data Clustering with Fuzzy C-Means: Problem, Reason, and Solution",

abstract = "Fuzzy C-Means (FCM) clustering algorithm is a popular unsupervised learning approach that has been extensively utilized in various domains. However, in this study, we point out a major problem faced by FCM when it is applied to the high-dimensional data, i.e., quite often the obtained prototypes (cluster centers) could not be distinguished with each other. Many studies have claimed that the concentration of the distance (CoD) could be a major reason for this phenomenon. This paper has therefore revisited this factor, and highlight that the CoD could not only lead to decreased performance, but sometimes also positively contribute to enhanced performance of the clustering algorithm. Instead, this paper point out the significance of features that are noisy and correlated, which could have a negative effect on FCM performance. Hence, to tackle the mentioned problem, we resort to a neural network model, i.e., the autoencoder, to reduce the dimensionality of the feature space while extracting features that are most informative. We conduct several experiments to show the validity of the proposed strategy for the FCM algorithm.",

keywords = "Fuzzy c-means, High-dimensional data, Autoencoder",

author = "Yinghua Shen and Hanyu E and Tianhua Chen and Zhi Xiao and Bingsheng Liu and Yuan Chen",

note = "Funding Information: This work was supported in part by the National Natural Science Foundation of China under Grant 72001032, Grant 72071021, Grant 72002152; in part by Natural Science Foundation of Chongqing under Grant cstc2020jcyj-bshX0013. Publisher Copyright: {\textcopyright} 2021, Springer Nature Switzerland AG. Copyright: Copyright 2021 Elsevier B.V., All rights reserved.; 16th International Work-Conference on Artificial Neural Networks, IWANN 2021 ; Conference date: 16-06-2021 Through 18-06-2021",

year = "2021",

month = sep,

day = "1",

doi = "10.1007/978-3-030-85030-2_8",

language = "English",

isbn = "9783030850296",

volume = "LNCS/LNTCS 12861",

series = "Lecture Notes in Computer Science",

publisher = "Springer Nature Switzerland AG",

pages = "89--100",

editor = "Ignacio Rojas and Gonzalo Joya and Andreu Catal{\`a}",

booktitle = "Advances in Computational Intelligence",

address = "Switzerland",

edition = "1st",

url = "http://iwann.uma.es/",

}

Shen, Y, E, H, Chen, T, Xiao, Z, Liu, B & Chen, Y 2021, High-dimensional Data Clustering with Fuzzy C-Means: Problem, Reason, and Solution. in I Rojas, G Joya & A Català (eds), Advances in Computational Intelligence: 16th International Work-Conference on Artificial Neural Networks, IWANN 2021, Virtual Event, June 16–18, 2021, Proceedings, Part I. 1st edn, vol. LNCS/LNTCS 12861, Lecture Notes in Computer Science, vol. LNCS/LNTCS 12861, Springer Nature Switzerland AG, Cham, pp. 89-100, 16th International Work-Conference on Artificial Neural Networks, Virtual, 16/06/21. https://doi.org/10.1007/978-3-030-85030-2_8

High-dimensional Data Clustering with Fuzzy C-Means : Problem, Reason, and Solution. / Shen, Yinghua; E, Hanyu; Chen, Tianhua et al.

Advances in Computational Intelligence: 16th International Work-Conference on Artificial Neural Networks, IWANN 2021, Virtual Event, June 16–18, 2021, Proceedings, Part I. ed. / Ignacio Rojas; Gonzalo Joya; Andreu Català. Vol. LNCS/LNTCS 12861 1st. ed. Cham : Springer Nature Switzerland AG, 2021. p. 89-100 (Lecture Notes in Computer Science; Vol. LNCS/LNTCS 12861).

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution › peer-review

TY - GEN

T1 - High-dimensional Data Clustering with Fuzzy C-Means

T2 - 16th International Work-Conference on Artificial Neural Networks

AU - Shen, Yinghua

AU - E, Hanyu

AU - Chen, Tianhua

AU - Xiao, Zhi

AU - Liu, Bingsheng

AU - Chen, Yuan

N1 - Conference code: 16

PY - 2021/9/1

Y1 - 2021/9/1

N2 - Fuzzy C-Means (FCM) clustering algorithm is a popular unsupervised learning approach that has been extensively utilized in various domains. However, in this study, we point out a major problem faced by FCM when it is applied to the high-dimensional data, i.e., quite often the obtained prototypes (cluster centers) could not be distinguished with each other. Many studies have claimed that the concentration of the distance (CoD) could be a major reason for this phenomenon. This paper has therefore revisited this factor, and highlight that the CoD could not only lead to decreased performance, but sometimes also positively contribute to enhanced performance of the clustering algorithm. Instead, this paper point out the significance of features that are noisy and correlated, which could have a negative effect on FCM performance. Hence, to tackle the mentioned problem, we resort to a neural network model, i.e., the autoencoder, to reduce the dimensionality of the feature space while extracting features that are most informative. We conduct several experiments to show the validity of the proposed strategy for the FCM algorithm.

AB - Fuzzy C-Means (FCM) clustering algorithm is a popular unsupervised learning approach that has been extensively utilized in various domains. However, in this study, we point out a major problem faced by FCM when it is applied to the high-dimensional data, i.e., quite often the obtained prototypes (cluster centers) could not be distinguished with each other. Many studies have claimed that the concentration of the distance (CoD) could be a major reason for this phenomenon. This paper has therefore revisited this factor, and highlight that the CoD could not only lead to decreased performance, but sometimes also positively contribute to enhanced performance of the clustering algorithm. Instead, this paper point out the significance of features that are noisy and correlated, which could have a negative effect on FCM performance. Hence, to tackle the mentioned problem, we resort to a neural network model, i.e., the autoencoder, to reduce the dimensionality of the feature space while extracting features that are most informative. We conduct several experiments to show the validity of the proposed strategy for the FCM algorithm.

KW - Fuzzy c-means

KW - High-dimensional data

KW - Autoencoder

UR - https://link.springer.com/book/10.1007/978-3-030-85030-2

UR - http://www.scopus.com/inward/record.url?scp=85115198549&partnerID=8YFLogxK

U2 - 10.1007/978-3-030-85030-2_8

DO - 10.1007/978-3-030-85030-2_8

M3 - Conference contribution

SN - 9783030850296

VL - LNCS/LNTCS 12861

T3 - Lecture Notes in Computer Science

SP - 89

EP - 100

BT - Advances in Computational Intelligence

A2 - Rojas, Ignacio

A2 - Joya, Gonzalo

A2 - Català, Andreu

PB - Springer Nature Switzerland AG

CY - Cham

Y2 - 16 June 2021 through 18 June 2021

ER -

Shen Y, E H, Chen T, Xiao Z, Liu B, Chen Y. High-dimensional Data Clustering with Fuzzy C-Means: Problem, Reason, and Solution. In Rojas I, Joya G, Català A, editors, Advances in Computational Intelligence: 16th International Work-Conference on Artificial Neural Networks, IWANN 2021, Virtual Event, June 16–18, 2021, Proceedings, Part I. 1st ed. Vol. LNCS/LNTCS 12861. Cham: Springer Nature Switzerland AG. 2021. p. 89-100. (Lecture Notes in Computer Science). Epub 2021 Aug 21. doi: 10.1007/978-3-030-85030-2_8

High-dimensional Data Clustering with Fuzzy C-Means: Problem, Reason, and Solution

Abstract

Publication series

Conference

Access to Document

Other files and links

Fingerprint

Cite this