Non-tuned machine learning approach for hydrological time series forecasting

Zaher Mundher Yaseen, Mohammed Falah Allawi, Ali A. Yousif, Othman Jaafar, Firdaus Mohamad Hamzah, Ahmed El-Shafie

Research output: Contribution to journalArticle

19 Citations (Scopus)

Abstract

Stream-flow forecasting is a crucial task for hydrological science. Throughout the literature, traditional and artificial intelligence models have been applied to this task. An attempt to explore and develop better expert models is an ongoing endeavor for this hydrological application. In addition, the accuracy of modeling, confidence and practicality of the model are the other significant problems that need to be considered. Accordingly, this study investigates modern non-tuned machine learning data-driven approach, namely extreme learning machine (ELM). This data-driven approach is containing single layer feedforward neural network that selects the input variables randomly and determine the output weights systematically. To demonstrate the reliability and the effectiveness, one-step-ahead stream-flow forecasting based on three time-scale pattern (daily, mean weekly and mean monthly) for Johor river, Malaysia, were implemented. Artificial neural network (ANN) model is used for comparison and evaluation. The results indicated ELM approach superior the ANN model level accuracies and time consuming in addition to precision forecasting in tropical zone. In measureable terms, the dominance of ELM model over ANN model was indicated in accordance with coefficient determination (R2) root-mean-square error (RMSE) and mean absolute error (MAE). The results were obtained for example the daily time scale R2 = 0.94 and 0.90, RMSE = 2.78 and 11.63, and MAE = 0.10 and 0.43, for ELM and ANN models respectively.

Original languageEnglish
Pages (from-to)1-13
Number of pages13
JournalNeural Computing and Applications
DOIs
Publication statusAccepted/In press - 16 Dec 2016

Fingerprint

Learning systems
Time series
Neural networks
Stream flow
Mean square error
Feedforward neural networks
Artificial intelligence
Rivers

Keywords

  • Artificial neural network
  • Extreme learning machine
  • Multiple time horizons
  • Stream-flow forecasting
  • Tropical environment

ASJC Scopus subject areas

  • Software
  • Artificial Intelligence

Cite this

Non-tuned machine learning approach for hydrological time series forecasting. / Yaseen, Zaher Mundher; Allawi, Mohammed Falah; Yousif, Ali A.; Jaafar, Othman; Mohamad Hamzah, Firdaus; El-Shafie, Ahmed.

In: Neural Computing and Applications, 16.12.2016, p. 1-13.

Research output: Contribution to journalArticle

Yaseen, Zaher Mundher ; Allawi, Mohammed Falah ; Yousif, Ali A. ; Jaafar, Othman ; Mohamad Hamzah, Firdaus ; El-Shafie, Ahmed. / Non-tuned machine learning approach for hydrological time series forecasting. In: Neural Computing and Applications. 2016 ; pp. 1-13.
@article{9c6e261d03e4411f9da7bd8f5dcc5a6f,
title = "Non-tuned machine learning approach for hydrological time series forecasting",
abstract = "Stream-flow forecasting is a crucial task for hydrological science. Throughout the literature, traditional and artificial intelligence models have been applied to this task. An attempt to explore and develop better expert models is an ongoing endeavor for this hydrological application. In addition, the accuracy of modeling, confidence and practicality of the model are the other significant problems that need to be considered. Accordingly, this study investigates modern non-tuned machine learning data-driven approach, namely extreme learning machine (ELM). This data-driven approach is containing single layer feedforward neural network that selects the input variables randomly and determine the output weights systematically. To demonstrate the reliability and the effectiveness, one-step-ahead stream-flow forecasting based on three time-scale pattern (daily, mean weekly and mean monthly) for Johor river, Malaysia, were implemented. Artificial neural network (ANN) model is used for comparison and evaluation. The results indicated ELM approach superior the ANN model level accuracies and time consuming in addition to precision forecasting in tropical zone. In measureable terms, the dominance of ELM model over ANN model was indicated in accordance with coefficient determination (R2) root-mean-square error (RMSE) and mean absolute error (MAE). The results were obtained for example the daily time scale R2 = 0.94 and 0.90, RMSE = 2.78 and 11.63, and MAE = 0.10 and 0.43, for ELM and ANN models respectively.",
keywords = "Artificial neural network, Extreme learning machine, Multiple time horizons, Stream-flow forecasting, Tropical environment",
author = "Yaseen, {Zaher Mundher} and Allawi, {Mohammed Falah} and Yousif, {Ali A.} and Othman Jaafar and {Mohamad Hamzah}, Firdaus and Ahmed El-Shafie",
year = "2016",
month = "12",
day = "16",
doi = "10.1007/s00521-016-2763-0",
language = "English",
pages = "1--13",
journal = "Neural Computing and Applications",
issn = "0941-0643",
publisher = "Springer London",

}

TY - JOUR

T1 - Non-tuned machine learning approach for hydrological time series forecasting

AU - Yaseen, Zaher Mundher

AU - Allawi, Mohammed Falah

AU - Yousif, Ali A.

AU - Jaafar, Othman

AU - Mohamad Hamzah, Firdaus

AU - El-Shafie, Ahmed

PY - 2016/12/16

Y1 - 2016/12/16

N2 - Stream-flow forecasting is a crucial task for hydrological science. Throughout the literature, traditional and artificial intelligence models have been applied to this task. An attempt to explore and develop better expert models is an ongoing endeavor for this hydrological application. In addition, the accuracy of modeling, confidence and practicality of the model are the other significant problems that need to be considered. Accordingly, this study investigates modern non-tuned machine learning data-driven approach, namely extreme learning machine (ELM). This data-driven approach is containing single layer feedforward neural network that selects the input variables randomly and determine the output weights systematically. To demonstrate the reliability and the effectiveness, one-step-ahead stream-flow forecasting based on three time-scale pattern (daily, mean weekly and mean monthly) for Johor river, Malaysia, were implemented. Artificial neural network (ANN) model is used for comparison and evaluation. The results indicated ELM approach superior the ANN model level accuracies and time consuming in addition to precision forecasting in tropical zone. In measureable terms, the dominance of ELM model over ANN model was indicated in accordance with coefficient determination (R2) root-mean-square error (RMSE) and mean absolute error (MAE). The results were obtained for example the daily time scale R2 = 0.94 and 0.90, RMSE = 2.78 and 11.63, and MAE = 0.10 and 0.43, for ELM and ANN models respectively.

AB - Stream-flow forecasting is a crucial task for hydrological science. Throughout the literature, traditional and artificial intelligence models have been applied to this task. An attempt to explore and develop better expert models is an ongoing endeavor for this hydrological application. In addition, the accuracy of modeling, confidence and practicality of the model are the other significant problems that need to be considered. Accordingly, this study investigates modern non-tuned machine learning data-driven approach, namely extreme learning machine (ELM). This data-driven approach is containing single layer feedforward neural network that selects the input variables randomly and determine the output weights systematically. To demonstrate the reliability and the effectiveness, one-step-ahead stream-flow forecasting based on three time-scale pattern (daily, mean weekly and mean monthly) for Johor river, Malaysia, were implemented. Artificial neural network (ANN) model is used for comparison and evaluation. The results indicated ELM approach superior the ANN model level accuracies and time consuming in addition to precision forecasting in tropical zone. In measureable terms, the dominance of ELM model over ANN model was indicated in accordance with coefficient determination (R2) root-mean-square error (RMSE) and mean absolute error (MAE). The results were obtained for example the daily time scale R2 = 0.94 and 0.90, RMSE = 2.78 and 11.63, and MAE = 0.10 and 0.43, for ELM and ANN models respectively.

KW - Artificial neural network

KW - Extreme learning machine

KW - Multiple time horizons

KW - Stream-flow forecasting

KW - Tropical environment

UR - http://www.scopus.com/inward/record.url?scp=85006356473&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85006356473&partnerID=8YFLogxK

U2 - 10.1007/s00521-016-2763-0

DO - 10.1007/s00521-016-2763-0

M3 - Article

SP - 1

EP - 13

JO - Neural Computing and Applications

JF - Neural Computing and Applications

SN - 0941-0643

ER -