Gradient-boosted trees often outperform linear models on tabular da…

Question

Gradient-boosted trees often outperform linear models on tabular data because:

Accepted Answer

They capture nonlinearities and feature interactions with less manual feature crafting — GBTs chew nonlinear interactions with less hand-featurization — still can overfit.

Answer

They require no data at all

Answer

They always extrapolate better outside training range

Answer

They never overfit

Gradient-boosted trees often outperform linear models on tabular data because:

Short Explanation

Full Explanation