Predicting Water Quality with Machine Learning

Abstract

Measuring water quality using machine learning methods is the main objective of this study. One way to measure the purity of water is by looking at its potability, which is a numerical expression. In order to determine the water's overall potability, this research used the following water quality criteria. The variables included turbidity, organic carbon, hardness, solids, chloromines, sulfate, conductivity, and trihalomethanes. These characteristics serve as a feature vector that represents the water quality. The study used Decision Tree (DT) and K-Nearest Neighbor (KNN) classification methods to assess the water quality class. A real dataset including data from several places in Andhra Pradesh and a parameter-generated synthetic dataset were both used in the experiments. The KNN classifier was shown to perform better than other classifiers based on the findings of two other kinds of classifiers. The results show that machine learning methods can reliably forecast the potability. Index keywords include topics such as classification, data mining, potability, and water quality parameters. Machine Learning, Supervised Learning, Decision Tree, Hyper Parameter Tuning, and Python are some of the terms used in this context. I.