Machine learning for water quality classification

Abstract

In the past years, there has been a lot of interest in water quality and its prediction as there are many pollutants that affect water quality. The techniques provided herein will help us in controlling and reducing the risks of water pollution. In this study, we will discuss concepts related to machine learning models and their applications for water quality classification (WQC). Three machine learning algorithms, J48, Naïve Bayes, and multi-layer perceptron (MLP), were used for WQC prediction. The dataset used contains 10 features, and in order to evaluate the machine's algorithms and their performance, some accuracy measurements were used. Our study showed that the proposed models can accurately classify water quality. By analyzing the results, it was found that the MLP algorithm achieved the highest accuracy for WQC prediction as compared to other algorithms. HIGHLIGHTS Machine learning concept was adopted to analyze the water quality.; The accuracy of multi-layer perceptron (MLP), is higher than other machine learning algorithms for water quality classification (WQC).; Extraction of useful and relevant features increases classification accuracy using principal component analysis (PCA).; The PCA was used for dimensionality reduction and extracts the most dominant water quality features.;

Keywords

• data mining
• machine learning
• multi-layer perceptron algorithm
• pearson's correlation coefficient
• principal component analysis (pca)
• water quality classification (wqc)