Why fit residuals reduce loss? (Or why fit the residuals of the weak model to get the residual tree and then linearly add the weak model to get the stronger model?)

NoSuchKey

Guess you like

Origin blog.csdn.net/weixin_40363423/article/details/97240551