A Machine Learning Approach To Predict the Quality of Drinkable Water From Different Source

Abstract

Machine learning is now a crucial technique for data analysis, classification, and prediction due to the exponential growth in the amount of data available on the aquatic environment. Data-driven models based on machine learning have the ability to effectively tackle more complicated nonlinear problems, in contrast to conventional models utilized in water-related research. Models and findings from machine learning have been used in water environment research to build, monitor, simulate, evaluate, and optimize various water treatment and management systems. Machine learning can also offer solutions for reducing water pollution, enhancing water quality, and managing the security of the watershed environment. In this paper, we explain the use of machine learning algorithms to assess the water quality in various water contexts, including surface water, groundwater, drinking water, sewage, and others. We also suggest potential future uses of machine learning techniques in aquatic contexts. For forecasting the potability of water, we employ the KNN, SVM, Random Forest, Decission Tree, and XGBoost algorithms. Pre-trained KNN algorithms were employed.