Hi,
This is more a theoretical question. I am trying to prove that the AdaBoost
algorithm is equivalent to forward stagewise additive modelling using an
exponential loss function.
The proof is actually in Page 305 of "The Elements of Statistical Learning"
by I cannot follow how it goes from eq. (10.11) to solve the optimization
problem for Beta.
Thanks for your help.
Cheers,
Leo Guelman
|