Water Potability Prediction Model Based on Machine Learning Techniques

Abstract

Earth is surrounded by 70% water of different qualities. Various pollutants have threatened water quality over the last few years. While conventional methods for monitoring water quality entail manually gathering water samples and analysing them in a lab, these procedures are sometimes time-consuming and expensive. Machine learning (ML) models can be used as a less expensive and more productive option to human labour to address these issues. These models are essential in reducing water pollution because they can accurately estimate the quality of water based on a number of significant characteristics. In order to accurately estimate water quality, the present study uses artificial intelligence (AI) techniques. It makes use of the PyCaret platform to find pertinent characteristics and the quadratic discriminant analysis (QDA) model to provide reliable findings. The dataset contains 9 parameters based on these parameters, the model finds whether a given sample of water is potable or not.