Ensemble learning combines the predictions of multiple models (weak learners) to create a single strong learner with better performance and stability.
The 'Wisdom of Crowds' applied to AI.
Dominates Kaggle competitions (e.g., XGBoost).