Quality Monitoring of Abu Dhabi Drinking Water Using Machine Learning Classifiers

Abstract

Supplying high quality drinking water is important for human health. It is essential to monitor the drinking water quality to prevent any health damage and avoid water pollution. This paper focuses on exploring the use of machine learning classifiers for monitoring drinking water quality in Abu Dhabi. Five machine learning algorithms (Logistic regression, a Support Vector Machines, a K-Nearest Neighbors, a Naive Bayes, and Decision Trees) were applied for water quality monitoring using standard water physical and chemical data as Abu Dhabi department of water requirements. The results show Decision Tree to be more efficient when compared with Logistic regression, a Support Vector Machines, a K-Nearest Neighbors, and Naive Bayes to predict water quality levels. Decision Tree showed 97.7011% accuracy, out-performing Logistic regression, a Support Vector Machines, a K-Nearest Neighbors, and Naive Bayes.