Differential evolution-based weighted combination  of distance metrics for k-means clustering

Fuad Muhammad Marwan Muhammad

doi:10.1007/978-3-319-13749-0

Differential evolution-based weighted combination of distance metrics for k-means clustering

Fuad Muhammad Marwan Muhammad

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

6 Citations (Scopus)

Abstract

Bio-inspired optimization algorithms have been successfully used to solve many problems in engineering, science, and economics. In computer science bio-inspired optimization has different applications in different domains such as software engineering, networks, data mining, and many others. One of the main tasks in data mining is clustering, namely k-means clustering. Distance metrics are at the heart of all data mining tasks. In this paper we present a new method which applies differential evolution, one of the main bio-inspired optimization algorithms, on a time series k-means clustering task to set the weights of the distance metrics used in a combination that is used to cluster the time series. The weights are obtained by applying an optimization process that gives optimal clustering quality. We show through extensive experiments how this optimized combination outperforms all the other stand-alone distance metrics, all by keeping the same low complexity of the distance metrics used in the combination.

Original language	English
Title of host publication	Theory and Practice of Natural Computing - 3rd International Conference, TPNC 2014, Proceedings
Editors	Adrian-Horia Dediu, Manuel Lozano, Carlos Martín-Vide
Publisher	Springer-Verlag Italia
Pages	193-204
Number of pages	12
ISBN (Electronic)	9783319137483
DOIs	https://doi.org/10.1007/978-3-319-13749-0
Publication status	Published - 1 Jan 2014
Event	3rd International Conference on the Theory and Practice of Natural Computing - Granada, Spain Duration: 9 Dec 2014 → 11 Dec 2014 Conference number: 3rd

Publication series

Name	Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume	8890
ISSN (Print)	0302-9743
ISSN (Electronic)	1611-3349

Conference

Conference	3rd International Conference on the Theory and Practice of Natural Computing
Abbreviated title	TPNC 2014
Country/Territory	Spain
City	Granada
Period	9/12/14 → 11/12/14

Keywords

Differential evolution
Distance metrics
Evolutionary computing
k-means clustering
Time series data mining

ASJC Scopus subject areas

Theoretical Computer Science
Computer Science(all)

Access to Document

10.1007/978-3-319-13749-0

Cite this

Muhammad Marwan Muhammad, F. (2014). Differential evolution-based weighted combination of distance metrics for k-means clustering. In A.-H. Dediu, M. Lozano, & C. Martín-Vide (Eds.), Theory and Practice of Natural Computing - 3rd International Conference, TPNC 2014, Proceedings (pp. 193-204). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 8890). Springer-Verlag Italia. https://doi.org/10.1007/978-3-319-13749-0

Differential evolution-based weighted combination of distance metrics for k-means clustering. / Muhammad Marwan Muhammad, Fuad.
Theory and Practice of Natural Computing - 3rd International Conference, TPNC 2014, Proceedings. ed. / Adrian-Horia Dediu; Manuel Lozano; Carlos Martín-Vide. Springer-Verlag Italia, 2014. p. 193-204 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 8890).

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

Muhammad Marwan Muhammad, F 2014, Differential evolution-based weighted combination of distance metrics for k-means clustering. in A-H Dediu, M Lozano & C Martín-Vide (eds), Theory and Practice of Natural Computing - 3rd International Conference, TPNC 2014, Proceedings. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 8890, Springer-Verlag Italia, pp. 193-204, 3rd International Conference on the Theory and Practice of Natural Computing, Granada, Spain, 9/12/14. https://doi.org/10.1007/978-3-319-13749-0

Muhammad Marwan Muhammad F. Differential evolution-based weighted combination of distance metrics for k-means clustering. In Dediu AH, Lozano M, Martín-Vide C, editors, Theory and Practice of Natural Computing - 3rd International Conference, TPNC 2014, Proceedings. Springer-Verlag Italia. 2014. p. 193-204. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)). doi: 10.1007/978-3-319-13749-0

Muhammad Marwan Muhammad, Fuad. / Differential evolution-based weighted combination of distance metrics for k-means clustering. Theory and Practice of Natural Computing - 3rd International Conference, TPNC 2014, Proceedings. editor / Adrian-Horia Dediu ; Manuel Lozano ; Carlos Martín-Vide. Springer-Verlag Italia, 2014. pp. 193-204 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).

@inproceedings{80d52f448417476bad2d6546991ecf96,

title = "Differential evolution-based weighted combination of distance metrics for k-means clustering",

abstract = "Bio-inspired optimization algorithms have been successfully used to solve many problems in engineering, science, and economics. In computer science bio-inspired optimization has different applications in different domains such as software engineering, networks, data mining, and many others. One of the main tasks in data mining is clustering, namely k-means clustering. Distance metrics are at the heart of all data mining tasks. In this paper we present a new method which applies differential evolution, one of the main bio-inspired optimization algorithms, on a time series k-means clustering task to set the weights of the distance metrics used in a combination that is used to cluster the time series. The weights are obtained by applying an optimization process that gives optimal clustering quality. We show through extensive experiments how this optimized combination outperforms all the other stand-alone distance metrics, all by keeping the same low complexity of the distance metrics used in the combination.",

keywords = "Differential evolution, Distance metrics, Evolutionary computing, k-means clustering, Time series data mining",

author = "{Muhammad Marwan Muhammad}, Fuad",

year = "2014",

month = jan,

day = "1",

doi = "10.1007/978-3-319-13749-0",

language = "English",

series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

publisher = "Springer-Verlag Italia",

pages = "193--204",

editor = "Adrian-Horia Dediu and Manuel Lozano and Carlos Mart{\'i}n-Vide",

booktitle = "Theory and Practice of Natural Computing - 3rd International Conference, TPNC 2014, Proceedings",

address = "Italy",

note = "3rd International Conference on the Theory and Practice of Natural Computing, TPNC 2014 ; Conference date: 09-12-2014 Through 11-12-2014",

}

TY - GEN

T1 - Differential evolution-based weighted combination of distance metrics for k-means clustering

AU - Muhammad Marwan Muhammad, Fuad

N1 - Conference code: 3rd

PY - 2014/1/1

Y1 - 2014/1/1

N2 - Bio-inspired optimization algorithms have been successfully used to solve many problems in engineering, science, and economics. In computer science bio-inspired optimization has different applications in different domains such as software engineering, networks, data mining, and many others. One of the main tasks in data mining is clustering, namely k-means clustering. Distance metrics are at the heart of all data mining tasks. In this paper we present a new method which applies differential evolution, one of the main bio-inspired optimization algorithms, on a time series k-means clustering task to set the weights of the distance metrics used in a combination that is used to cluster the time series. The weights are obtained by applying an optimization process that gives optimal clustering quality. We show through extensive experiments how this optimized combination outperforms all the other stand-alone distance metrics, all by keeping the same low complexity of the distance metrics used in the combination.

AB - Bio-inspired optimization algorithms have been successfully used to solve many problems in engineering, science, and economics. In computer science bio-inspired optimization has different applications in different domains such as software engineering, networks, data mining, and many others. One of the main tasks in data mining is clustering, namely k-means clustering. Distance metrics are at the heart of all data mining tasks. In this paper we present a new method which applies differential evolution, one of the main bio-inspired optimization algorithms, on a time series k-means clustering task to set the weights of the distance metrics used in a combination that is used to cluster the time series. The weights are obtained by applying an optimization process that gives optimal clustering quality. We show through extensive experiments how this optimized combination outperforms all the other stand-alone distance metrics, all by keeping the same low complexity of the distance metrics used in the combination.

KW - Differential evolution

KW - Distance metrics

KW - Evolutionary computing

KW - k-means clustering

KW - Time series data mining

UR - http://www.scopus.com/inward/record.url?scp=84917710778&partnerID=8YFLogxK

U2 - 10.1007/978-3-319-13749-0

DO - 10.1007/978-3-319-13749-0

M3 - Conference proceeding

AN - SCOPUS:84917710778

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

SP - 193

EP - 204

BT - Theory and Practice of Natural Computing - 3rd International Conference, TPNC 2014, Proceedings

A2 - Dediu, Adrian-Horia

A2 - Lozano, Manuel

A2 - Martín-Vide, Carlos

PB - Springer-Verlag Italia

T2 - 3rd International Conference on the Theory and Practice of Natural Computing

Y2 - 9 December 2014 through 11 December 2014

ER -