Gradient boosting
Gradient boosting: An ensemble method sequentially adding weak models to correct errors according to a loss gradient. Small trees and shrinkage control the process. It is effective on tabular data but sensitive to hyperparameters, leakage and temporal evaluation. This perspective connects the definition with practical behaviour in artificial intelligence systems.
In italiano: Gradient BoostingLegge III · Come una rete impara dai datiSmall trees and shrinkage control the process. It is effective on tabular data but sensitive to hyperparameters, leakage and temporal evaluation.