Integrating feature extraction approaches with hybrid emotional neural networks for water quality index modeling

The establishment of water quality prediction models is vital for aquatic ecosystems analysis. traditional methods index (WQI) analysis are time-consuming and associated with a high degree errors. These days, the application artificial intelligence (AI) based trending capturing nonlinear complex processes. Therefore, present study was conducted to predict WQI in Kinta River, Malaysia by employing hybrid AI model i.e., GA-EANN (genetic algorithm-emotional neural network). extreme gradient boosting (XGB) neuro-sensitivity (NSA) approaches were utilized feature extraction, six different combinations derived examine relationship among (WQ) variables. efficacy proposed evaluated against backpropagation network (BPNN) multilinear regression (MLR) during calibration, validation periods on Nash–Sutcliffe efficiency (NSE), mean square error (MSE), root (RMSE), absolute percentage (MAPE), correlation coefficient (CC) indicators. According results appraisal produced better outcomes (NSE = 0.9233/ 0.9018, MSE 10.5195/ 9.7889 mg/L, RMSE 3.2434/ 3.1287 MAPE 3.8032/ 3.0348 CC 0.9609/ 0.9496) calibration/ phases than BPNN MLR models. In addition, indicate performance suitability five input parameters predicting site.