Machine learning-based prediction of drinking water quality index in Western Tehran using KAN, MLP, and traditional models.

In this study, the water quality index (WQI) was calculated using multivariate statistics, incorporating physical, chemical, and microbiological analysis of water samples taken from water supply networks in the western district of Tehran from 2021 to 2024. The principal drinking water parameters such as pH, total hardness, turbidity, lead (Pb), chloride (Cl), fluoride (F), total dissolved solids (TDS), sulfate (SO), nitrate (NO), nitrite (NO), calcium (Ca), magnesium (Mg), arsenic (As), mercury (Hg), cadmium (Cd), fecal coliform and total residual chlorine (Ch) were selected according to Iranian national water standards. The WQI index was predicted using various machine learning algorithms, including multiple linear regression (MLR), support vector machine (SVM) regression, extreme gradient boosting (XGBoost), Random Forest (RF) regression, multilayer perceptron (MLP), and Kolmogorov-Arnold networks (KAN). The results showed that the traditionally computed WQI values ranged from 75 to 86, indicating that water quality varied from fair to good for drinking purposes. The prepared WQI maps revealed that the water quality of 71% of the evaluated areas between 2021 and 2022 was in the good range, whereas the water quality of 50% and 87% of the locations was rated as good in the subsequent years. Moreover, excellent prediction results were obtained, with an R of 0.901 and root mean squared error (RMSE) of 0.286 for the MLP model, and an R of 0.9527 and RMSE of 0.197 for the KAN model. The results obtained indicate that both the KAN and MLP models are effective for accurately predicting the WQI.