Fuzzy gaussian potential neural networks using a functional reasoning

Mohammad Teshnehlab, Keigo Watanabe

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

1 Citation (Scopus)

Abstract

This paper presents the principal design of a fuzzy gaussian potential neural network (FGPNN) to achieve high capability to learn expert control rules of the fuzzy controller. In this construction, each membership function consists of a gaussian potential function (GPF) which causes the utilization of a reduced number of labels, and eventually the complexity of structural design becomes simple, specially for large scale inputs, This in turn reduces the learning trials, to improve the learning speed. Thus, the time of the training process, which is based on the ba~k-propagation method, is shortened. The construction of an FGPNN is carried out with the minimum number of GPF, based on the number of input patterns, to learn the mean vectors and shapes of the individual GPFs that basically depend on the desired trajectory. Finally, we provide a simulation to evaluate the proposed method for a multi input-output, twolink manipulator.

Original languageEnglish
Title of host publicationLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
PublisherSpringer Verlag
Pages34-47
Number of pages14
Volume1011
ISBN (Print)9783540606079
Publication statusPublished - 1995
Externally publishedYes
Event3rd World Wisepersons Workshop on Fuzzy Logic and Neural Networks/Genetic Algorithms, WWW 1994 - Nagoya, Japan
Duration: Aug 9 1994Aug 10 1994

Publication series

NameLecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
Volume1011
ISSN (Print)03029743
ISSN (Electronic)16113349

Other

Other3rd World Wisepersons Workshop on Fuzzy Logic and Neural Networks/Genetic Algorithms, WWW 1994
CountryJapan
CityNagoya
Period8/9/948/10/94

Fingerprint

Gaussian Function
Reasoning
Neural Networks
Potential Function
Neural networks
Membership functions
Structural design
Manipulators
Labels
Structural Design
Trajectories
Manipulator
Fuzzy Controller
Membership Function
Controllers
Trajectory
Propagation
Evaluate
Output
Simulation

ASJC Scopus subject areas

  • Computer Science(all)
  • Theoretical Computer Science

Cite this

Teshnehlab, M., & Watanabe, K. (1995). Fuzzy gaussian potential neural networks using a functional reasoning. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) (Vol. 1011, pp. 34-47). (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 1011). Springer Verlag.

Fuzzy gaussian potential neural networks using a functional reasoning. / Teshnehlab, Mohammad; Watanabe, Keigo.

Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 1011 Springer Verlag, 1995. p. 34-47 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics); Vol. 1011).

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

Teshnehlab, M & Watanabe, K 1995, Fuzzy gaussian potential neural networks using a functional reasoning. in Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). vol. 1011, Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics), vol. 1011, Springer Verlag, pp. 34-47, 3rd World Wisepersons Workshop on Fuzzy Logic and Neural Networks/Genetic Algorithms, WWW 1994, Nagoya, Japan, 8/9/94.
Teshnehlab M, Watanabe K. Fuzzy gaussian potential neural networks using a functional reasoning. In Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 1011. Springer Verlag. 1995. p. 34-47. (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
Teshnehlab, Mohammad ; Watanabe, Keigo. / Fuzzy gaussian potential neural networks using a functional reasoning. Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics). Vol. 1011 Springer Verlag, 1995. pp. 34-47 (Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)).
@inproceedings{32ccc0763d144a94a9af0dbc4e480fa2,
title = "Fuzzy gaussian potential neural networks using a functional reasoning",
abstract = "This paper presents the principal design of a fuzzy gaussian potential neural network (FGPNN) to achieve high capability to learn expert control rules of the fuzzy controller. In this construction, each membership function consists of a gaussian potential function (GPF) which causes the utilization of a reduced number of labels, and eventually the complexity of structural design becomes simple, specially for large scale inputs, This in turn reduces the learning trials, to improve the learning speed. Thus, the time of the training process, which is based on the ba~k-propagation method, is shortened. The construction of an FGPNN is carried out with the minimum number of GPF, based on the number of input patterns, to learn the mean vectors and shapes of the individual GPFs that basically depend on the desired trajectory. Finally, we provide a simulation to evaluate the proposed method for a multi input-output, twolink manipulator.",
author = "Mohammad Teshnehlab and Keigo Watanabe",
year = "1995",
language = "English",
isbn = "9783540606079",
volume = "1011",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "34--47",
booktitle = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",

}

TY - GEN

T1 - Fuzzy gaussian potential neural networks using a functional reasoning

AU - Teshnehlab, Mohammad

AU - Watanabe, Keigo

PY - 1995

Y1 - 1995

N2 - This paper presents the principal design of a fuzzy gaussian potential neural network (FGPNN) to achieve high capability to learn expert control rules of the fuzzy controller. In this construction, each membership function consists of a gaussian potential function (GPF) which causes the utilization of a reduced number of labels, and eventually the complexity of structural design becomes simple, specially for large scale inputs, This in turn reduces the learning trials, to improve the learning speed. Thus, the time of the training process, which is based on the ba~k-propagation method, is shortened. The construction of an FGPNN is carried out with the minimum number of GPF, based on the number of input patterns, to learn the mean vectors and shapes of the individual GPFs that basically depend on the desired trajectory. Finally, we provide a simulation to evaluate the proposed method for a multi input-output, twolink manipulator.

AB - This paper presents the principal design of a fuzzy gaussian potential neural network (FGPNN) to achieve high capability to learn expert control rules of the fuzzy controller. In this construction, each membership function consists of a gaussian potential function (GPF) which causes the utilization of a reduced number of labels, and eventually the complexity of structural design becomes simple, specially for large scale inputs, This in turn reduces the learning trials, to improve the learning speed. Thus, the time of the training process, which is based on the ba~k-propagation method, is shortened. The construction of an FGPNN is carried out with the minimum number of GPF, based on the number of input patterns, to learn the mean vectors and shapes of the individual GPFs that basically depend on the desired trajectory. Finally, we provide a simulation to evaluate the proposed method for a multi input-output, twolink manipulator.

UR - http://www.scopus.com/inward/record.url?scp=84948137734&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84948137734&partnerID=8YFLogxK

M3 - Conference contribution

SN - 9783540606079

VL - 1011

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

SP - 34

EP - 47

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

PB - Springer Verlag

ER -