Water Quality Prediction Using Gradient Boosting Machines: A Case Study of the Cauvery and Tamirabarani Rivers

This study employs datasets from Cauvery, Tamirabarani rivers for 2019-2023 as a basis for analyzing machine learning techniques for water quality prediction with an important focus on Gradient Boosting Machine. In GBM research, the model is validated according with its use in the accurate identification of water quality trends with respect to dissolved oxygen rate <tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$(\text{mg} / \mathrm{L}), \text{pH}, \text{BOD}(\text{mg} / \mathrm{L})$</tex>, fecal coliform (MPN/100ml), Conductivity (mmhos/cm), Nitrate (mg/L), Turbidity (NTU), Chloride, Total Alkalinity (mg/L), Chloride (mg/L), COD (mg/L), Ammonia (mg/L), Hardness (mg/L), Calcium (mg/L), Magnesium (mg/L), Sodium (mg/L), Total Dissolved Solids (<tex xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">$\text{mg} / \mathrm{L}$</tex>) and the aim is to provide enhanced control for polluting agents and water resources. The study evaluates whether the model performs well in identifying enhancements or deterioration of water quality, or stability with the help of these indices. Further a portal has also been developed which also provides basic feature to enter and get feedback on current water quality parameters in real time with reference to trends in past data. Therefore, there is evidence that validates the efficacy of GBM for ecological approaches in connection with resource management programs of monitoring plans, in terms of the strong predictive utility observed.