Optimasi Hyperparameter Random Forest untuk Prediksi Kualitas Air

Abstract

Water is an essential source of life for all living beings, making the preservation of water quality crucial for human health and the sustainability of ecosystems. However, population growth, industrial development, and other economic activities have led to a decline in water quality and an increase in pollution. To address this issue, early detection through the application of data mining techniques, particularly classification, is necessary to predict the quality of consumable water. This study aims to enhance the accuracy of water quality prediction by applying hyperparameter optimization techniques using the Grid Search method on the Random Forest algorithm. The results show that hyperparameter optimization improved water quality prediction accuracy from 88.33% to 91.32%. This improvement underscores the effectiveness of machine learning techniques in monitoring water quality and contributes to better decision-making to safeguard water resources and protect public health. Furthermore, this research provides insights into the importance of data mining techniques in identifying relevant patterns in water quality data, thereby helping to prevent health risks associated with contaminated water.