Water Quality analysis of Well Water using Machine Learning Techniques

Abstract

Water quality is continuously deteriorating with the release of unprocessed industrial effluents, sewage, and wastewater from the households, agriculture runoff, and untreated wastewater has contaminated the water bodies like rivers, lakes, and ponds which in turn affects the groundwater. The quality of water is being affected by several parameters such as pollution, acid rain, and other chemicals from agriculture runoff which include fertilizers and pesticides which make the water toxic. The quality of water that is being taken has a direct effect on the health of a living organism, the consumption of impure water causes various water-borne diseases like cholera, diarrhea and affects child mortality. To overcome these problems, in this project we are going to predict the water quality using various machine learning(ML) algorithms. The training phase includes the usage of various models such as Logistic Regressor (LR), Random Forest(RF), Extra Tree, Decision Tree(DT), Support Vector Machine (SVM), and XG Boost. The models were evaluated and the results of five machine learning models were compared. Out of these five models, Random Forest performed best with prediction accuracy of 98% and precision of 97%.