Boosting

Regularization in Boosting

Set of techniques (shrinkage, subsampling, L1/L2 penalties) controlling the complexity of the boosting model to prevent overfitting and improve generalization.

← Wstecz