A Sustainability Model for Water Quality Classification Using Machine Learning Techniques

Water quality affects both the environmental system and the human health. Water is utilized for different types of purposes, such as drinking, agriculture, and industry. Water is an essential part of several Sustainable Development Goals (SDGs) established by the United Nations. The classification of water quality is therefore crucial for maintaining a healthy community and has grown to play an important role in preserving the aquatic ecosystem. The water quality is measured by calculating and utilizing the weighted arithmetic index strategy. This paper presents the applicable of six machine learning algorithms on water quality that are K-nearest neighbor (KNN), Random Forest (RF), Decision Tree (DT), Adaptive Boosting (AdaBoost), Logistic Regression (LR), Support Vector Machine (SVM). The experiment uses one of the most popular datasets of water quality, which is called the classification of water quality. The experimental dataset is the water quality classification dataset from a website called Kaggle. The dataset includes 21 columns with 8000 records that are split into 70% training set and 30% testing set. The experimental results present the tested accuracy of 86.9%, 98.5%, 98.7%, 98.6%, 77.1%, and 77.6% using K-Nearest Neighbor, Random Forest, Decision Tree, Adaptive Boosting, Logistic Regression, and Support Vector Machine respectively. The decision tree algorithm is illustrated the best with the highest accuracy.