News for hackers. Stuff that matters.
Re:Decision Trees – Gradient Boosting Models by brendamorgan on Monday July 20, 2015 @04:14AM (#9914622)
Gradient Boosting models are another variant of ensemble models, different from Random Forest. In Random Forest (RF) models, goal is to build many-many overfitted models each on subset of training data, and combine their individual prediction to make final prediction. In Gradient Boosting models, goal is to build series of many-many underfitted models, each bettering errors of previous model, and cumulative prediction is used to make final prediction.

There are no comments on this comment yet.