What is error decomposition in ensemble learning?

Error decomposition breaks ensemble error into bias, variance, and noise components.

In ensemble learning, prediction error can be decomposed into bias (systematic error), variance (sensitivity to training data), and irreducible noise. Bagging reduces variance by averaging multiple models trained on different samples. Boosting reduces bias by sequentially correcting errors. Stacking can reduce both by learning optimal combinations. Understanding error decomposition helps choose the right ensemble strategy.