TY  - JOUR
TT  - Performance comparison of multilayer perceptron and wavelet-multilayer perceptron models in terms of daily streamflow prediction
AU  - Altunkaynak, Abdüsselam
AU  - Nıgussıe, Tewodros Assefa
PY  - 2016
DA  - December
JF  - Dicle Üniversitesi Mühendislik Fakültesi Mühendislik Dergisi
JO  - DUJE
PB  - Dicle Üniversitesi
WT  - DergiPark
SN  - 1309-8640
SP  - 291
EP  - 308
VL  - 7
IS  - 2
KW  - Akım
KW  - Ayrık Dalgacık Dönüşümü
KW  - Çok Tabakalı Perseptron
KW  - Tahmin
N2  - Accurate streamflow prediction is important for sustainable water resources management. Direct use of observed data in developing prediction models has resulted in inaccuracies and predictions are with short lead times. DWT was used to decompose the observed data into components with the objective of enhancing the prediction accuracy and prediction lead times. The decomposed data were used as input into multilayer perceptron (MP) to develop a new approach for predicting daily streamflow for lead times up to 7 days. The new approach was called Wavelet-Multilayer Perceptron (W-MP). Twelve years of approved daily streamflow data were obtained from Station 02231000, USA. Seven years of data were used for calibration and the remaining 5 years of data were used for prediction. The new approach was compared to the stand-alone MP model by taking root mean squared error, coefficient of efficiency and skill score into consideration. The results showed that the W-MP model performed better than the stand-alone MP model and the prediction accuracy increased with the use of decomposed signals up to prediction lead time of 4 days. This indicates that the W-MP model can predict daily streamflow better than MP with extended lead time.
UR  - https://dergipark.org.tr/tr/pub/dumf/issue//312768
L1  - https://dergipark.org.tr/tr/download/article-file/302791
ER  -