Spatial Prediction of Groundwater Potentiality in Large Semi-Arid and Karstic Mountainous Region Using Machine Learning Models

The drinking and irrigation water scarcity is a major global issue, particularly in arid semi-arid zones. In rural areas, groundwater could be used as an alternative additional supply source order to reduce human suffering terms of scarcity. this context, the purpose present study facilitate potentiality mapping via spatial-modelling techniques, individual ensemble machine-learning models. Random forest (RF), logistic regression (LR), decision tree (DT) artificial neural networks (ANNs) are main algorithms study. preparation maps was assembled into 11 ensembles Overall, about 374 springs identified inventoried mountain area. spring inventory data randomly divided training (75%) testing (25%) datasets. Twenty-four influencing factors (GIFs) were selected based on multicollinearity test information gain calculation. results validated using statistical measures receiver operating characteristic curve (ROC) method. Finally, ranking 15 models achieved with prioritization rank method compound factor (CF) most stable suitable for mountainous aquifers compared success prediction rate. efficient model area under validation RF-LR-DT-ANN Moreover, indicate that best RF-DT RF-LR-DT