I checked Mease and Wyner's paper Boosted Classification Trees and
Class Probability/Quantile Estimation.
In that paper, the formula on page 13 seems different from the one in
hw4. But their formula makes more sense to me. I guess there are still
some typo in hw4.
On Mar 10, 10:25 pm, "karlmzh...@gmail.com" <karlmzh...@gmail.com>
wrote: