Determination of the complexity fitted model structure of Radial Basis Function Neural Networks

A. Várkonyi-Kóczy, Balazs Tusor, Adrienn Dineva

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

Abstract

One of the disadvantages of using Artificial Neural Networks (ANNs) is their significant training time need, which scales with the complexity of the network and with the complexity of the problem that is needed to be solved. Radial Basis Function Neural Networks (RBFNNs) are neural networks that use the linear combination of radial basis functions, utilizing hybrid learning procedures which can solve the time requirement problem efficiently. However, it is not trivial to determine their structural parameters, such as the number of neurons as well as the parameters of each neuron. To solve that problem we have developed a new training method: we apply a clustering step to the training data, which results in information both about the quasi-optimum number of necessary neurons in the model and the approximate parameters of the neurons.

Original languageEnglish
Title of host publicationINES 2013 - IEEE 17th International Conference on Intelligent Engineering Systems, Proceedings
Pages237-242
Number of pages6
DOIs
Publication statusPublished - 2013
Event17th IEEE International Conference on Intelligent Engineering Systems, INES 2013 - San Jose, Costa Rica
Duration: Jun 19 2013Jun 21 2013

Other

Other17th IEEE International Conference on Intelligent Engineering Systems, INES 2013
CountryCosta Rica
CitySan Jose
Period6/19/136/21/13

Fingerprint

Model structures
Neurons
Neural networks

ASJC Scopus subject areas

  • Artificial Intelligence
  • Control and Systems Engineering

Cite this

Várkonyi-Kóczy, A., Tusor, B., & Dineva, A. (2013). Determination of the complexity fitted model structure of Radial Basis Function Neural Networks. In INES 2013 - IEEE 17th International Conference on Intelligent Engineering Systems, Proceedings (pp. 237-242). [6632818] https://doi.org/10.1109/INES.2013.6632818

Determination of the complexity fitted model structure of Radial Basis Function Neural Networks. / Várkonyi-Kóczy, A.; Tusor, Balazs; Dineva, Adrienn.

INES 2013 - IEEE 17th International Conference on Intelligent Engineering Systems, Proceedings. 2013. p. 237-242 6632818.

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

Várkonyi-Kóczy, A, Tusor, B & Dineva, A 2013, Determination of the complexity fitted model structure of Radial Basis Function Neural Networks. in INES 2013 - IEEE 17th International Conference on Intelligent Engineering Systems, Proceedings., 6632818, pp. 237-242, 17th IEEE International Conference on Intelligent Engineering Systems, INES 2013, San Jose, Costa Rica, 6/19/13. https://doi.org/10.1109/INES.2013.6632818
Várkonyi-Kóczy A, Tusor B, Dineva A. Determination of the complexity fitted model structure of Radial Basis Function Neural Networks. In INES 2013 - IEEE 17th International Conference on Intelligent Engineering Systems, Proceedings. 2013. p. 237-242. 6632818 https://doi.org/10.1109/INES.2013.6632818
Várkonyi-Kóczy, A. ; Tusor, Balazs ; Dineva, Adrienn. / Determination of the complexity fitted model structure of Radial Basis Function Neural Networks. INES 2013 - IEEE 17th International Conference on Intelligent Engineering Systems, Proceedings. 2013. pp. 237-242
@inproceedings{348cdb5f119142939e0a41b16fbdd482,
title = "Determination of the complexity fitted model structure of Radial Basis Function Neural Networks",
abstract = "One of the disadvantages of using Artificial Neural Networks (ANNs) is their significant training time need, which scales with the complexity of the network and with the complexity of the problem that is needed to be solved. Radial Basis Function Neural Networks (RBFNNs) are neural networks that use the linear combination of radial basis functions, utilizing hybrid learning procedures which can solve the time requirement problem efficiently. However, it is not trivial to determine their structural parameters, such as the number of neurons as well as the parameters of each neuron. To solve that problem we have developed a new training method: we apply a clustering step to the training data, which results in information both about the quasi-optimum number of necessary neurons in the model and the approximate parameters of the neurons.",
author = "A. V{\'a}rkonyi-K{\'o}czy and Balazs Tusor and Adrienn Dineva",
year = "2013",
doi = "10.1109/INES.2013.6632818",
language = "English",
isbn = "9781479908288",
pages = "237--242",
booktitle = "INES 2013 - IEEE 17th International Conference on Intelligent Engineering Systems, Proceedings",

}

TY - GEN

T1 - Determination of the complexity fitted model structure of Radial Basis Function Neural Networks

AU - Várkonyi-Kóczy, A.

AU - Tusor, Balazs

AU - Dineva, Adrienn

PY - 2013

Y1 - 2013

N2 - One of the disadvantages of using Artificial Neural Networks (ANNs) is their significant training time need, which scales with the complexity of the network and with the complexity of the problem that is needed to be solved. Radial Basis Function Neural Networks (RBFNNs) are neural networks that use the linear combination of radial basis functions, utilizing hybrid learning procedures which can solve the time requirement problem efficiently. However, it is not trivial to determine their structural parameters, such as the number of neurons as well as the parameters of each neuron. To solve that problem we have developed a new training method: we apply a clustering step to the training data, which results in information both about the quasi-optimum number of necessary neurons in the model and the approximate parameters of the neurons.

AB - One of the disadvantages of using Artificial Neural Networks (ANNs) is their significant training time need, which scales with the complexity of the network and with the complexity of the problem that is needed to be solved. Radial Basis Function Neural Networks (RBFNNs) are neural networks that use the linear combination of radial basis functions, utilizing hybrid learning procedures which can solve the time requirement problem efficiently. However, it is not trivial to determine their structural parameters, such as the number of neurons as well as the parameters of each neuron. To solve that problem we have developed a new training method: we apply a clustering step to the training data, which results in information both about the quasi-optimum number of necessary neurons in the model and the approximate parameters of the neurons.

UR - http://www.scopus.com/inward/record.url?scp=84889673522&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=84889673522&partnerID=8YFLogxK

U2 - 10.1109/INES.2013.6632818

DO - 10.1109/INES.2013.6632818

M3 - Conference contribution

AN - SCOPUS:84889673522

SN - 9781479908288

SP - 237

EP - 242

BT - INES 2013 - IEEE 17th International Conference on Intelligent Engineering Systems, Proceedings

ER -