Application of Computational Model Based Probabilistic Neural Network for Surface Water Quality Prediction

Applications of artificial intelligence (AI) models have been massively explored for various engineering and sciences domains over the past two decades. Their capacity in modeling complex problems confirmed motivated researchers to explore their merit different disciplines. The use AI-models (probabilistic neural network multilayer perceptron network) estimation water quality indicators (namely dissolved oxygen (DO) five days biochemical demand (BOD5)) were reported this study. WQ parameters based on four input modelling scenarios was adopted. Monthly data duration from January 2006 December 2015 used as building prediction model. proposed established utilizing many physical chemical variables, such turbidity, calcium (Ca), pH, temperature (T), total solids (TDS), Sulfate (SO4), suspended (TSS), alkalinity variables. evaluated performance using statistical metrics evaluation results showed that terms accuracy increases with addition more variables some cases. performances PNN model superior MLPNN both DO BOD parameters. study concluded is a good tool estimating optimal predicting are (R2 = 0.93, RMSE 0.231 MAE 0.197). best Do 0.94, 0.222 0.175).