A technique that combines predictions from multiple machine learning models to create a stronger, more robust predictor. Ensemble methods like bagging, boosting, and stacking reduce overfitting, improve generalization, and often achieve better performance than individual models alone.
