Machine Learning for Water Resources Management: Comparative study

  Aniss MOUMEN, National School of Applied Sciences Kenitra, Morocco
  Nezha MEJJAD, Faculty of Sciences, Morocco
  Rachid EL ANSARI, National School of Applied Sciences Kenitra, Morocco
  Mohamed EL BOUHADDIOUI, National school of Mines Rabat, Morocco

Machine learning techniques present an alternative way to estimate, evaluate and predict the evolution of water parameters. Indeed, the decider makers consider all this information to improve the situation of water resources. In this paper, we present a meta-analysis of a systematic literature review about applications of Machine learning to address issues related to water resources. We have found more than 160 references. To elaborate this literature review we used scientific databases: Scopus, ScienceDirect and IEEE. After applying word cloud requests, we have identified many topics: Water quality, Groundwater, Water Surface,  Irrigation, drought. Also, the most recurrent machine learning technics used by researchers in this field are Neural Network, Deep Learning, SVM, Regressions, Xgboost, ELM. To conclude this paper, we present a comparative discussion between machine learning technics according to their metrics to evaluate the quality of those models.

Keywords: Machine learning|SLR|Water ressources|Metrics

A105156AM