Performance of Different Ensemble Machine Learning Algorithms in Classification of the Water Quality Indices
Central Asian Journal of Water Research, cilt.12, sa.1, ss.68-92, 2026 (Scopus)
- Yayın Türü: Makale / Tam Makale
- Cilt numarası: 12 Sayı: 1
- Basım Tarihi: 2026
- Doi Numarası: 10.29258/cajwr/2026-r1.v12-1/68-92.eng
- Dergi Adı: Central Asian Journal of Water Research
- Derginin Tarandığı İndeksler: Scopus
- Sayfa Sayıları: ss.68-92
- Anahtar Kelimeler: Classification, Ensemble Learning, Feature selection, Machine learning, Water Quality Index
- Gazi Üniversitesi Adresli: Evet
Özet
Monitoring the quality of ground freshwater resources is crucial due to their limited availability and susceptibility to contamination from unchecked human activities. The Water Quality Index (WQI) serves as a tool for assessing basin water quality, yet its complexity and time-consuming nature have steered scientists toward simpler methods like machine learning algorithms. This study evaluated four machine learning algorithms using an ensemble learning approach to construct a high-performing classification model. Additionally, three feature selection methods using machine learning were applied to determine the most influential water quality parameters in the classification process. Among the classifiers, XGBoost demonstrated exceptional performance, achieving 96.9696% accuracy when considering all parameters. To optimize cost-effectiveness, the XGBoost classifier with a 95.606% accuracy using 10-Fold Cross Validation and seven key parameters, identified by Backward Feature Elimination, is recommended for classification on the dataset used in this study.