Please use this identifier to cite or link to this item:
|Title:||Methods to improve neural network performance in daily flows prediction|
|Authors:||Wu, C. L.|
Li, Y. S.
|Subjects:||Daily flows prediction|
Artificial neural network
Singular spectral analysis
Wavelet multi-resolution analysis
|Source:||Journal of hydrology, 15 June 2009, v. 372, no. 1-4, p. 80-93.|
|Abstract:||In this paper, three data-preprocessing techniques, moving average (MA), singular spectrum analysis (SSA), and wavelet multi-resolution analysis (WMRA), were coupled with artificial neural network (ANN) to improve the estimate of daily flows. Six models, including the original ANN model without data preprocessing, were set up and evaluated. Five new models were ANN-MA, ANN-SSA1, ANN-SSA2, ANN-WMRA1, and ANN-WMRA2. The ANN-MA was derived from the raw ANN model combined with the MA. The ANN-SSA1, ANN-SSA2, ANN-WMRA1 and ANN-WMRA2 were generated by using the original ANN model coupled with SSA and WMRA in terms of two different means. Two daily flow series from different watersheds in China (Lushui and Daning) were used in six models for three prediction horizons (i.e., 1-, 2-, and 3-day-ahead forecast). The poor performance on ANN forecast models was mainly due to the existence of the lagged prediction. The ANN-MA, among six models, performed best and eradicated the lag effect. The performances from the ANN-SSA1 and ANN-SSA2 were similar, and the performances from the ANN-WMRA1 and ANN-WMRA2 were also similar. However, the models based on the SSA presented better performance than the models based on the WMRA at all forecast horizons, which meant that the SSA is more effective than the WMRA in improving the ANN performance in the current study. Based on an overall consideration including the model performance and the complexity of modeling, the ANN-MA model was optimal, then the ANN model coupled with SSA, and finally the ANN model coupled with WMRA.|
|Rights:||Journal of Hydrology © 2009 Elsevier B.V. The journal web site is located at http://www.sciencedirect.com.|
|Appears in Collections:||CEE Journal/Magazine Articles|
Files in This Item:
|JH7.pdf||Pre-published version||668.07 kB||Adobe PDF||View/Open|
All items in the PolyU Institutional Repository are protected by copyright, with all rights reserved, unless otherwise indicated. No item in the PolyU IR may be reproduced for commercial or resale purposes.