Question regarding Gentle Boost(GB)

[Arvind]

Hi,

I have questions regarding the working of gentle adaboost. After doing
some research I understood that GB differs from AB in how
i) the weak classifiers give their o/p[probability measure and not a
binary output]
ii) How the samples are reassigned weights?

But I fail to understand the equation governing this. Can some one
provide an example and explain how the weak classifier o/p is diff
from adaboost and how the weights are reassigned.Thanks for ur help.