Method evidence record
Gradient Boosting
Gradient Boosting is an ensemble learning method, formalised by Jerome H. Friedman in 2001, that combines a sequence of weak learners — typically shallow decision trees — so that each new tree is fitted to minimise the residual errors of the trees before it. It is the core algorithm behind popular implementations such as XGBoost, LightGBM and CatBoost.
Source record
Citations copied verbatim from the method’s source record. No claim-level verification is inferred from them.
Gradient Boosting Machine (Friedman's Gradient Boosting)
Taxonomic method record · ml-model / machine-learning
Open full method Curated claims
Claims persisted in the evidence ledger, each with its own assessment.
No curated claims yet
This view does not invent a claim assessment when the ledger has none.
Related methods
Generated from the method graph and shown as machine-suggested relations — no evidence claim is inferred.