DocumentCode :
1841967
Title :
A novel approach for training neural networks for long-term prediction
Author :
Hashem, S. ; Ashour, Z.H. ; Abdel Gawad, E.F. ; Hakeem, A. Abdel
Author_Institution :
Dept. of Eng. Math. & Phys., Cairo Univ., Giza, Egypt
Volume :
3
fYear :
1999
fDate :
1999
Firstpage :
1594
Abstract :
Neural networks have been widely used in performing time series prediction. Long-term prediction is generally far more difficult than short-term prediction, because of the difficulty in modeling the system dynamics far ahead. In this paper, we present a novel approach for training neural networks to perform long-term prediction. Our approach relies on the utilization of traditional time series analysis, based on Box-Jenkins methodology (1976), to: (1) determine the appropriate neural network architecture, (2) select the inputs to the neural network, and (3) determine the appropriate lead time for updating the connection-weights of the neural network during training. We demonstrate the effectiveness of this approach in producing accurate multistep ahead prediction on some real-world problems as well as on simulated time series data
Keywords :
forecasting theory; iterative methods; learning (artificial intelligence); neural net architecture; time series; Box-Jenkins methodology; connection-weights; lead time; long-term prediction; multistep ahead prediction; neural network architecture; neural network training; simulated time series data; system dynamics modeling; time series analysis; time series prediction; Autoregressive processes; Economic forecasting; Mathematics; Network topology; Neural networks; Performance evaluation; Physics; Predictive models; Recurrent neural networks; Time series analysis;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Neural Networks, 1999. IJCNN '99. International Joint Conference on
Conference_Location :
Washington, DC
ISSN :
1098-7576
Print_ISBN :
0-7803-5529-6
Type :
conf
DOI :
10.1109/IJCNN.1999.832609
Filename :
832609
Link To Document :
بازگشت