Water Quality Assessment of the River Cauvery: Prediction of Treatment Recommendations Using LR-GB Stacking Ensemble Model

Abstract

Machine learning techniques are increasingly being applied in environmental protection, particularly for analyzing the quality of air, water and soil. One of the most important problems in current environment is water pollution and it’s our responsibility to protect one of our primordial elements of the earth. Many researchers are currently concentrating in preventing pollution by developing advanced monitoring systems, sustainable treatment methods and data-driven predictive models for early detection and control. The water quality parameter values were used to determine the contamination level and it helps to decide the treatment recommendations for the Cauvery river dataset.  In this work, the LR-GB Stacking ensemble model was proposed and efficiently used to predict the future treatment recommendations for the Cauvery river stations. This model achieved higher prediction accuracy, precision and recall when compared with the other machine learning models.