Machine Learning Models for Predicting Fish Mortality in Polluted Rivers

The paper presents a novel approach to predicting fish mortality in polluted river ecosystems using advanced machine learning techniques. By correlating real-time water pollution data with fish health parameters such as hemoglobin levels, enzyme activity, and oxidative stress markers, we developed accurate models (Random Forest, SVM, ANN) capable of identifying mortality risks. This study contributes to environmental monitoring, decision-making, and sustainable aquatic resource management.