Rival penalized Self-Organizing Map

Lap Tak Law; Yiu Ming Cheung

Rival penalized Self-Organizing Map

Lap Tak Law^*, Yiu Ming Cheung

^*Corresponding author for this work

Department of Computer Science

Research output: Chapter in book/report/conference proceeding › Conference proceeding › peer-review

1 Citation (Scopus)

Abstract

Kohonen's Self-Organizing Map (SOM) is one of the most commonly used competitive learning algorithms that provide a topological mapping from the input space to the output space. In the conventional SOM, it needs to choose an appropriate learning rate as well as a monotonically decreasing function that lowers the learning rate with time to ensure the convergence of the map. Otherwise, its performance may seriously deteriorate. In this paper, we therefore propose a novel Rival Penalized Self-Organizing Map (RPSOM) learning algorithm, which dynamically penalizes a set of rivals towards driving far away from the input data set during the learning. Compared to the existing methods, this new one need not select the monotonically decreasing function of the learning rate, but still gives a robust result. The experiments have shown its outstanding performance in comparison with the existing algorithms.

Original language	English
Title of host publication	Proceedings of the IASTED International Conference on Neural Networks and Computational Intelligence 2004
Publisher	International Conference on Neural Networks and Computational Intelligence
Pages	142-145
Number of pages	4
Publication status	Published - Feb 2004
Event	IASTED International Conference on Neural Networks and Computational Intelligence 2004 - Grindelwald, Switzerland Duration: 23 Feb 2004 → 25 Feb 2004

Publication series

Name	Proceedings of the IASTED International Conference on Neural Networks and Computational Intelligence

Conference

Conference	IASTED International Conference on Neural Networks and Computational Intelligence 2004
Country/Territory	Switzerland
City	Grindelwald
Period	23/02/04 → 25/02/04

Scopus Subject Areas

General Engineering

User-Defined Keywords

Rival Penalization Controlled Competitive Learning
Rival Penalized Self-Organizing Map
Self-Orgainzing Map

Cite this

Law, L. T., & Cheung, Y. M. (2004). Rival penalized Self-Organizing Map. In Proceedings of the IASTED International Conference on Neural Networks and Computational Intelligence 2004 (pp. 142-145). (Proceedings of the IASTED International Conference on Neural Networks and Computational Intelligence). International Conference on Neural Networks and Computational Intelligence.

Law, Lap Tak ; Cheung, Yiu Ming. / Rival penalized Self-Organizing Map. Proceedings of the IASTED International Conference on Neural Networks and Computational Intelligence 2004. International Conference on Neural Networks and Computational Intelligence, 2004. pp. 142-145 (Proceedings of the IASTED International Conference on Neural Networks and Computational Intelligence).

@inproceedings{31dcd052bff34929b6b564f1a56ab8f9,

title = "Rival penalized Self-Organizing Map",

abstract = "Kohonen's Self-Organizing Map (SOM) is one of the most commonly used competitive learning algorithms that provide a topological mapping from the input space to the output space. In the conventional SOM, it needs to choose an appropriate learning rate as well as a monotonically decreasing function that lowers the learning rate with time to ensure the convergence of the map. Otherwise, its performance may seriously deteriorate. In this paper, we therefore propose a novel Rival Penalized Self-Organizing Map (RPSOM) learning algorithm, which dynamically penalizes a set of rivals towards driving far away from the input data set during the learning. Compared to the existing methods, this new one need not select the monotonically decreasing function of the learning rate, but still gives a robust result. The experiments have shown its outstanding performance in comparison with the existing algorithms.",

keywords = "Rival Penalization Controlled Competitive Learning, Rival Penalized Self-Organizing Map, Self-Orgainzing Map",

author = "Law, {Lap Tak} and Cheung, {Yiu Ming}",

note = "Copyright: Copyright 2008 Elsevier B.V., All rights reserved.; IASTED International Conference on Neural Networks and Computational Intelligence 2004 ; Conference date: 23-02-2004 Through 25-02-2004",

year = "2004",

month = feb,

language = "English",

series = "Proceedings of the IASTED International Conference on Neural Networks and Computational Intelligence",

publisher = "International Conference on Neural Networks and Computational Intelligence",

pages = "142--145",

booktitle = "Proceedings of the IASTED International Conference on Neural Networks and Computational Intelligence 2004",

}

Law, LT & Cheung, YM 2004, Rival penalized Self-Organizing Map. in Proceedings of the IASTED International Conference on Neural Networks and Computational Intelligence 2004. Proceedings of the IASTED International Conference on Neural Networks and Computational Intelligence, International Conference on Neural Networks and Computational Intelligence, pp. 142-145, IASTED International Conference on Neural Networks and Computational Intelligence 2004, Grindelwald, Switzerland, 23/02/04.

Rival penalized Self-Organizing Map. / Law, Lap Tak; Cheung, Yiu Ming.
Proceedings of the IASTED International Conference on Neural Networks and Computational Intelligence 2004. International Conference on Neural Networks and Computational Intelligence, 2004. p. 142-145 (Proceedings of the IASTED International Conference on Neural Networks and Computational Intelligence).

Research output: Chapter in book/report/conference proceeding › Conference proceeding › peer-review

TY - GEN

T1 - Rival penalized Self-Organizing Map

AU - Law, Lap Tak

AU - Cheung, Yiu Ming

PY - 2004/2

Y1 - 2004/2

N2 - Kohonen's Self-Organizing Map (SOM) is one of the most commonly used competitive learning algorithms that provide a topological mapping from the input space to the output space. In the conventional SOM, it needs to choose an appropriate learning rate as well as a monotonically decreasing function that lowers the learning rate with time to ensure the convergence of the map. Otherwise, its performance may seriously deteriorate. In this paper, we therefore propose a novel Rival Penalized Self-Organizing Map (RPSOM) learning algorithm, which dynamically penalizes a set of rivals towards driving far away from the input data set during the learning. Compared to the existing methods, this new one need not select the monotonically decreasing function of the learning rate, but still gives a robust result. The experiments have shown its outstanding performance in comparison with the existing algorithms.

AB - Kohonen's Self-Organizing Map (SOM) is one of the most commonly used competitive learning algorithms that provide a topological mapping from the input space to the output space. In the conventional SOM, it needs to choose an appropriate learning rate as well as a monotonically decreasing function that lowers the learning rate with time to ensure the convergence of the map. Otherwise, its performance may seriously deteriorate. In this paper, we therefore propose a novel Rival Penalized Self-Organizing Map (RPSOM) learning algorithm, which dynamically penalizes a set of rivals towards driving far away from the input data set during the learning. Compared to the existing methods, this new one need not select the monotonically decreasing function of the learning rate, but still gives a robust result. The experiments have shown its outstanding performance in comparison with the existing algorithms.

KW - Rival Penalization Controlled Competitive Learning

KW - Rival Penalized Self-Organizing Map

KW - Self-Orgainzing Map

UR - http://www.scopus.com/inward/record.url?scp=11144342160&partnerID=8YFLogxK

M3 - Conference proceeding

AN - SCOPUS:11144342160

T3 - Proceedings of the IASTED International Conference on Neural Networks and Computational Intelligence

SP - 142

EP - 145

BT - Proceedings of the IASTED International Conference on Neural Networks and Computational Intelligence 2004

PB - International Conference on Neural Networks and Computational Intelligence

T2 - IASTED International Conference on Neural Networks and Computational Intelligence 2004

Y2 - 23 February 2004 through 25 February 2004

ER -

Rival penalized Self-Organizing Map

Abstract

Publication series

Conference

Scopus Subject Areas

User-Defined Keywords

Other files and links

Cite this