Yale University
Department of Statistics
Seminar

Monday, February 19, 2001


Vladimir I. Koltchinskii
Department of Mathematics and Statistics
University of New Mexico

Probabilistic bounds for AdaBoost and other  learning algorithms

We present new probabilistic bounds on the generalization error of complex classifiers that are "combinations" of simpler classifiers from a base class of functions. The bounds in question are in terms of so called "margins" of combined classifiers and they apply to voting methods of combining classifiers.  Although the bounds of this type originated in computer science literature, their true nature is related to the inequalities developed in the theory of empirical processes and in Probability in Banach spaces, especially, to concentration inequalities of Talagrand and also to symmetrization and comparison inequalities.
 



 

Seminar to be held in Room 107, 24 Hillhouse Avenue at 4:15 pm