Work

Optimizing Cerebral Stroke Prediction: Combining SMOTE-ENN and XGBoost on Imbalanced Medical Data