Bagging (Bootstrap Aggregating)

An ensemble method that trains multiple versions of a model on randomly sampled subsets of the training data (with replacement) and averages their predictions to reduce variance.