The use of PCA and ANN to improve evaluation of the WQIclassic, development of a new index, and prediction of WQI, Coastel Constantinois, northern coast of eastern Algeria

Abstract The objective of this research was to arrive at a better assessment the quality surface water in Constantine region. focus on comparison three classical indices WQINSF (National Sanitation Foundation Water Quality Index), WQICCME (Canadian Council Ministers Environment Index) and WQIAP (weighted arithmetical development new index prediction by ANN (artificial neural network) WQI indices. principal components analysis (PCA) allows selection 10 parameters be used calculation WQI, eight as input for proposed (regularized WQI). However is applied search models developed WQI. results show that assesses better, regularized further promotes quality. WQIR shows that, after pollution peak, does not return its initial state. modeling approach offers an effective alternative predict it subsequently appears predicts WQIRregularized (R2 = 0.999) than classic model 0.99).