Savoga

Ensemble Methods


Ensemble methods are good to reduce variance.

There are several types of ensemble methods:

Bagging (Random Forests)

Bagging methods reduce variance and handle overfitting.

Predictors are chosen independently.

Boosting (AdaBoost, Gradient Boosting)

Boosting methods have the advantage of reducing the variance and the bias. However, these algorithms can overfit.

Predictors are chosen sequentially: predictors learn from the mistakes of the previous ones. The choice of the stopping criteria is critical to prevent overfitting.