Tree-based machine learning model to predict the performance of heavy metal removal by sulfate-reducing bacteria

N.R. Srinivasan, T.S. Anju, Anish Antony,
Tree-based machine learning model to predict the performance of heavy metal removal by sulfate-reducing bacteria,
Bioresource Technology Reports,
Volume 32,
2025,
102362,
ISSN 2589-014X,
https://doi.org/10.1016/j.biteb.2025.102362.
(https://www.sciencedirect.com/science/article/pii/S2589014X25003457)
Abstract: The present study uses tree-based machine learning models to predict how eight input features affect the removal of heavy metals by sulfate-reducing bacteria (SRB). The tree-based algorithms used in this work are CatBoost, RandomForest, AdaBoost, GradientBoosting, XGBoost, and LightGBM. The performance of the model is evaluated based on the value of the correlation coefficient (R2). Catboost algorithm shows the highest heavy metal removal efficiency. (Cd: CatBoost: Test R2 = 0.9126, Cu: Test R2 = 0.9068 compared to other algorithms. Therefore, these two algorithms can be used for small data points (〈1000) to predict the performance of a system that uses SRB to remove heavy metals in wastewater. Further, the scientific insights behind the performance are presented in this work along with experimental evidence.
Keywords: Machine learning; Heavy metal; Sulfate-reducing bacteria