Decision Tree and Support Vector Machine for Anomaly Detection in Water Distribution Networks

Abstract

Drinking water quality monitoring is essential these days as the available water may be affected by pollution and can cause several diseases. Hence, it's necessary to prevent any intrusion into water distribution systems and to detect pollution momentarily. To resolve concerns on intrusion detection we have various machine learning algorithms for classification but choosing the best one is an important task. For selecting the best algorithm for our water quality monitoring system, we conducted an experimental study on machine learning algorithms. In this experimental study, we analyzed the performance of the famous classification algorithms in the literature namely Decision Tree and Support Vector Machines using a real dataset retrieved from a Tunisian water treatment station.