Title :
Application of a Hybrid Algorithm in the Modular Neural Nets Trainning with Multilayers Specialists in Electric Disturbance Classification
Author :
Magalhães, R.M. ; Santos, C.K.S. ; Melo, J.D. ; de Medeiros, M.F. ; Neto, A. D Dória
Author_Institution :
Intell. Syst. Lab., Fed. Univ. of Rio Grande do Norte, Natal
Abstract :
This work evaluates the effectiveness of a new hybrid algorithm in the training of modular neural nets with multilayers specialists. The algorithm is applied on the classification of electric disturbances previously characterized and processed, making possible the comparison between its performance and multilayers perceptrons nets. It presents the algorithm´s development,such as the data acquisition methodology and its preprocessing.
Keywords :
learning (artificial intelligence); multilayer perceptrons; pattern classification; power system analysis computing; power system faults; data acquisition; data preprocessing; electric disturbance classification; hybrid algorithm; modular neural net training; multilayer perceptron; Backpropagation algorithms; Data acquisition; Discrete wavelet transforms; Jacobian matrices; Machine learning; Machine learning algorithms; Multi-layer neural network; Multilayer perceptrons; Neural networks; Neurons;
Conference_Titel :
Intelligent Engineering Systems, 2008. INES 2008. International Conference on
Conference_Location :
Miami, FL
Print_ISBN :
978-1-4244-2082-7
Electronic_ISBN :
978-1-4244-2083-4
DOI :
10.1109/INES.2008.4481280
