Prediction of Water Quality through Dissolved Oxygen Saturation using Data Mining: A Case Study of Puebla Mexico

Abstract

Computational sciences have been highlighted in the application to any area with good results. One of the primary interests of humanity is water quality because it is a vital resource for the existence of living beings. This research applies a data mining model based on CRISP-DM methodology built to classify water contamination in lotic systems in Puebla City, located in the center of Mexico. The classification was carried out through physicochemical parameters using a water quality evaluation model based on the amount of dissolved oxygen percentage (%DO). Results demonstrate that the application of decision trees and K-NN algorithms, using chemical parameters, are effective in determining the presence of contamination in lotic water bodies and represent a novel way to evaluate water quality in the water system along rivers in Puebla, Mexico.