Development and application of hybrid artificial intelligence models for groundwater potential mapping and assessment

Groundwater potential assessment is essential for optimum utilization and recharge of groundwater resources the proper development management an area. The main aim this study to develop accurate map Dak Nong Province (Vietnam) using hybrid artificial intelligence models, which are a combination Random Forest (RF) its Ensemble Framework (AdaBoost - ABRF, Bagging BRF LogitBoost LBRF). In study, twelve conditioning factors, namely topography (aspect, elevation, Topographic Wetness Index TWI, slope, curvature), hydrology (infiltration river density, rainfall, Sediment Transport STI, Stream Power SPI), land use, soil were used models. Well, yield data was also utilized validate zones.
 One Rule (R) feature selection method prioritize importance affecting parameters. results indicated that Average Merit (AM) rainfall factor highest (68.039), density lowest (53,969). Performance evaluation ML models done standard statistical indicators, including Area Under Receiver Operating Characteristic (ROC) curve (AUC). showed all four performed well in training (AUC ≥ 0.967) testing 0.734) phases, but performance ABRF (AUC=0.992) model best phase, whereas LBRF phase (AUC=0.776). present would be helpful sustainable development.