Title :
An efficient algorithm for identifying the structure of artificial neural networks for forecasting problems
Author :
Hegazy, Y.G. ; Salama, M.M.A.
Author_Institution :
Dept. of Electr. & Comput. Eng., Waterloo Univ., Ont., Canada
Abstract :
This paper presents a practical method for identifying the most suitable structure of back-propagation neural networks in forecasting problems. The method as based on processing the data with different time series analysis models in order to find out the principal components of the data that should be used as input variables. In order to demonstrate the effectiveness of the proposed method a practical case study is presented in this paper. The results of this study show how the proposed method is promising in forecasting problems
Keywords :
backpropagation; forecasting theory; neural nets; time series; algorithm; artificial neural networks; back-propagation; data processing; forecasting; time series analysis models; Artificial neural networks; Computer networks; Economic forecasting; Input variables; Load forecasting; Neural networks; Power system economics; Predictive models; Time series analysis; Yttrium;
Conference_Titel :
Circuits and Systems, 1994., Proceedings of the 37th Midwest Symposium on
Conference_Location :
Lafayette, LA
Print_ISBN :
0-7803-2428-5
DOI :
10.1109/MWSCAS.1994.519371
