Machine Learning Models for Predicting Anemia: Evaluation and Performance Insights

Anemia, characterized by insufficient levels of red blood cells or hemoglobin in the bloodstream, results in reduced oxygen delivery to tissues and manifests symptoms such as fatigue, weakness, pale skin, and shortness of breath, underscoring the critical importance of early detection and effective treatment. The application of machine learning (ML) models—including Random Forest, ExtraTrees, Gradient Boosting, XGBoost, and a Hybrid model—in predicting anemia is explored. Diagnostic features like hemoglobin levels with red blood cell counts were utilized from a diverse patient dataset, with rigorous preprocessing ensuring data quality and consistency. Each model underwent evaluation using standard metrics such as accuracy, precision, recall, and F1 score. Results highlight the Hybrid model as achieving top performance, with 90.83 % accuracy, 95.93 % precision, 88.74% recall, and an impressive F1 score of 92.17%. By integrating diverse ML algorithms, this approach enhances predictive accuracy, demonstrating the potential of advanced ML techniques to refine diagnostic precision and support clinical decision-making for optimal anemia management. Future research aims to optimize model architectures and validate their efficacy across broader patient populations to further enhance healthcare outcomes.