A Theory of Probabilistic Boosting, Decision Trees and Matryoshki

Etienne Grossmann

We present a theory of boosting probabilistic classifiers. We place ourselves in the situation of a user who only provides a stopping parameter and a probabilistic weak learner/cla…