Bias variance tradeoff boosting (xgboost) vs random forest (randomized bagging) which to use when?


I was looking up differences between boosting an bagging and I see this quoted everywhere

If the classifier is unstable (high variance), then we should apply Bagging. If the classifier is stable and simple (high bias) then we should apply Boosting.


Breiman [1996a] showed that Bagging is effective on ``unstable'' learning algorithms where small changes in the training set result in large changes in predictions. Breiman [1996a] claimed that neural networks and decision trees are examples of unstable learning algorithms.check last lines

Aren't gbdt's the most preferred/recommended learners in xgboost ? Which is contradictory considering how decision trees are apparently unstable learners.


Posted 2020-10-27T16:19:23.117

Reputation: 1



The decision trees used in gradient boosting are typically shallow decision trees (with only a few nodes). Limiting the depth or number of nodes in the decision tree makes them simple. This is different from fully developed decision trees used as standalone models.


Posted 2020-10-27T16:19:23.117

Reputation: 5 477

... and in Random Forests, too. – desertnaut – 2020-10-27T23:17:53.590