Water Quality Classification Using Ensembled Learning Methodology with Matlab Software for Kshipra River

Abstract

Large-scale pollution of water resources is one of the principal consequences of industrialization and urban expansion. Like pollution can be classified depending on factors such as pH, specific conductivity, dissolved oxygen (DO), temperature, and so on. The data acquired from the above attributes can aid in estimating the water quality of any body of water, although the prediction accuracy of traditional models is not particularly good due to unpredictability and imprecision. The ensemble modeling technique was used in this study to estimate the performance of the ANN study to predict the water quality of river Kshipra. The Artificial Neural Network (ANN) is a sophisticated data-driven model capable of capturing and representing both linear and non-linear correlations between input and output data. A total of 17 features is considered for determining the water quality. Some of these are: Temperature, pH, DO, COD, and Specific conductivity. After assessing the obtained results, the assembled ANN model was found to be a better classification model with an accuracy of 92.60%.