All terms
Foundations
Gradient Boosting
An ensemble method that builds models in sequence, each fixing the previous one's mistakes.
Definition
Gradient boosting builds a model out of many small models — usually decision trees — added one at a time, where each new tree focuses on correcting the errors of the ones before it. The result is a highly accurate predictor, and tools like XGBoost and LightGBM make it a top choice for spreadsheet-style data.