<![CDATA[This study evaluates the performance of ensemble-based machine learning models for early-stage diabetes prediction. Three classifiers Random Forest, XGBoost, and LightGBM were assessed under baseline and hyperparameter-tuned configurations using an 80–20 train–test split. Model performance was measured using accuracy, precision, recall, and F1-score. The results show that all models achieved high predictive performance, with test accuracy reaching up to 99.04%. Random Forest demonstrated stable and consistent results without significant improvement after tuning. XGBoost showed performance enhancement after hyperparameter optimization, improving its generalization ability. LightGBM achieved competitive baseline performance but experienced a slight decrease after tuning. Learning curve analysis indicates that all models benefit from increased training data, with reduced overfitting as dataset size grows. Overall, Random Forest and tuned XGBoost emerged as the most reliable models for early-stage diabetes prediction, demonstrating strong generalization and high classification accuracy.]]>
Copyrights © 2026
