Enhanced Water Quality Classification through Integrated Machine Learning and Ensemble Methods

With the continuous rise in the pollution level of drinking resources, water quality is becoming a major global concern. Thus, techniques for regulating and categorizing water according to the presence of impurities must be developed. In this research article, Machine Learning (ML) Supervised Gradient Boosting (GB) Technique namely XGBoost, CatBoost and LGBoost is used for the classification analysis. To further increase the Efficiency and robustness of the classification model, weighted voting and stacking ensemble techniques are utilized. The comparison of the accuracy of the classification models used are combined using the ensemble techniques and the new classifier achieved an accuracy of 94.70% which is higher as compared to the other classifiers used. Moreover, the Jaccard Index Comparison of all the algorithms shows the similarity of the values predicted and tested is higher for the Stacking Ensemble Algorithm with a value of 66%. This classification is essential in making smart decisions about sustainable water resource management.