Translator Disclaimer
22 March 1996 Neural networks adaptive wavelets for predictions of the northeastern Brazil monthly rainfall anomalies time series
Author Affiliations +
Neural networks were used to predict the anomalies of the time series of monthly rainfall of the Northeastern Region of Brazil. The forecasts made using a feedforward network with backpropagation algorithm from the original data were not satisfactory. We have therefore tried to combine two advanced methods, Wavelet Transform and Neural networks. Three more types of neural networks were used. The selected neural networks include the Time Delay Neural Networks (TDNN), Radial Basis Functions network and Neural Network Adaptive Wavelet. All networks were implemented in neural network simulator SNNS. The Neural Network Adaptive Wavelet was implemented by changing the standard sigmoidal nonlinearities to wavelet nonlinearities in the neurons. We compare the results obtained with unfiltered and filtered data. Using data obtained by filtering the wavelet transform coefficients significantly improved the results for all networks. The combination of TDNN with wavelet filtered data gave the best results.
© (1996) COPYRIGHT Society of Photo-Optical Instrumentation Engineers (SPIE). Downloading of the abstract is permitted for personal use only.
Weigang Li, Leonardo Deane Sa, G. S.S. Prasad, A. G. Nowosad, Mauricio Bolzan, and E. S. M. Chiang "Neural networks adaptive wavelets for predictions of the northeastern Brazil monthly rainfall anomalies time series", Proc. SPIE 2760, Applications and Science of Artificial Neural Networks II, (22 March 1996);


Audio signal compression using circular wavelet packets
Proceedings of SPIE (October 10 1994)
Denoising via adaptive lifting schemes
Proceedings of SPIE (December 03 2000)
Directional complex-wavelet processing
Proceedings of SPIE (December 03 2000)
Mass flow rate measurement of oil water two phase flow...
Proceedings of SPIE (October 12 2008)
Fuzzy recognition of noncompact musical objects
Proceedings of SPIE (March 26 1997)
A two stage neural network based method for cycle slip...
Proceedings of SPIE (April 07 2009)

Back to Top