Skip to content

Latest commit

 

History

History
18 lines (9 loc) · 357 Bytes

File metadata and controls

18 lines (9 loc) · 357 Bytes

Boosting

The general principal behind boosting is to train an ensemble of models sequentially, each model

Training

  • Train an ensemble of simple models sequentially, each model "focusing on" (and therefore improving upon) the mispredictions of the previous model.

Prediction/Inference *

AdaBoost

Gradient Boosting

CatBoost