Water Quality Prediction: Machine Learning Approaches for Sustainable Resource Management

Human being, wildlife, plants, processes, ecosystems, and other aspects all have water efficacy at their core. Quality of water has reduced in the last decades due to contamination and pollution. That is why people have gradually started understanding the imperative necessity of water protection and balance in the environment. The integration of AI and ML could generate great potential in boosting the classification and the prediction of status of water. This paper compares different ML approaches to analyse long-term WQ data and generate a method for an accurate future water quality. This investigation discusses the strength and weakness in each of the technique's performance that is important to design a hybrid system that has the performance, resilience, accurately and real time applicability of each of the individual method. This suggests that this issue is timely significantly because another line of approach towards avoiding the lethal efficiency of poor-quality water exists. The evaluated classifiers included Support Vector Machine, Random Forest, Logistic Regression, Decision Tree, CAT Boost, XG Boost, and Multilayer Perceptron with better statistics for the real-world modelling. To assess the effectiveness of the classification approaches, five evaluation metrics are being calculated which includes the measures such as accuracy, recall, precision, Pearson Correlation Coefficient, F1 score.