ASSESSMENT OF THE CHANGE IN THE NUMBER OF NEURONS IN HIDDEN LAYERS OF NEURAL NETWORKS FOR FAULT IDENTIFICATION IN ELECTRICAL SYSTEMS

David Targueta da Silva, Pedro Henrique Gouvêa Coelho, Joaquim Augusto Pinto Rodrigues, Luiz Biondi Neto

2010

Abstract

This work describes performance evaluation of ANNs (Artificial Neural Networks) used to identify faults in electrical systems for several number of neurons in the hidden layers. The number of neurons in the hidden layers depends on the complexity of the problem to be represented. Currently, there are no reliable rules for determining, a priori, the number of hidden neurons, so that such number depends largely on the experience of the practitioners who are designing the ANNs. This paper reports experiment results using neural networks varying the number of hidden neurons to aid the neural network user to find an adequate configuration in terms of number of neurons in the hidden layers so that ANNs be more efficient particularly for fault identification applications.

References

  1. Biondi, L. N., 1997. Hybrid System for Decision Support for Detection and Fault Diagnostics in Electrical Networks. M.Sc. Dissertation, PUC-Rio, Rio de Janeiro, R.J., Brazil. In Portuguese.
  2. Silva, D. T., 2008. Decision Support System for Fault Identification in Electrical Energy Distribution Systems Networks Using Artificial Neural Networks. M. Sc. Dissertation, State University of Rio de Janeiro, R.J., Brazil. In Portuguese.
  3. Ostafe, D., 2005. Neural Network Hidden Layer Number Determination Using Pattern Recognition Techniques. In Proceedings of the 2nd Romanian-Hungarian Joint Symposium on Applied Computational Intelligence (SACI 2005).
  4. Xu, S., and Chen, L., 2008. A Novel Approach for Determining the Optimal Number of Hidden Layer Neurons for FNN's and Its Application in Data Mining. In Proceedings of the 5th International Conference on Information Technology and Applications (ICITA 2008).
Download


Paper Citation


in Harvard Style

Targueta da Silva D., Henrique Gouvêa Coelho P., Augusto Pinto Rodrigues J. and Biondi Neto L. (2010). ASSESSMENT OF THE CHANGE IN THE NUMBER OF NEURONS IN HIDDEN LAYERS OF NEURAL NETWORKS FOR FAULT IDENTIFICATION IN ELECTRICAL SYSTEMS . In Proceedings of the 12th International Conference on Enterprise Information Systems - Volume 2: ICEIS, ISBN 978-989-8425-05-8, pages 309-313. DOI: 10.5220/0002910203090313


in Bibtex Style

@conference{iceis10,
author={David Targueta da Silva and Pedro Henrique Gouvêa Coelho and Joaquim Augusto Pinto Rodrigues and Luiz Biondi Neto},
title={ASSESSMENT OF THE CHANGE IN THE NUMBER OF NEURONS IN HIDDEN LAYERS OF NEURAL NETWORKS FOR FAULT IDENTIFICATION IN ELECTRICAL SYSTEMS },
booktitle={Proceedings of the 12th International Conference on Enterprise Information Systems - Volume 2: ICEIS,},
year={2010},
pages={309-313},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0002910203090313},
isbn={978-989-8425-05-8},
}


in EndNote Style

TY - CONF
JO - Proceedings of the 12th International Conference on Enterprise Information Systems - Volume 2: ICEIS,
TI - ASSESSMENT OF THE CHANGE IN THE NUMBER OF NEURONS IN HIDDEN LAYERS OF NEURAL NETWORKS FOR FAULT IDENTIFICATION IN ELECTRICAL SYSTEMS
SN - 978-989-8425-05-8
AU - Targueta da Silva D.
AU - Henrique Gouvêa Coelho P.
AU - Augusto Pinto Rodrigues J.
AU - Biondi Neto L.
PY - 2010
SP - 309
EP - 313
DO - 10.5220/0002910203090313