We present a generalized PAC bound for averaging classifiers which applies to base hypotheses with a bounded real valued output. In addition, we discuss several methods for quantitatively tightening the bound. In the process, a tightened version of the PAC-Bayes bound is proved.

15 pages

[email protected], www.cs.cmu.edu/~jcl
*Institute for Adaptive and Neural Computation, University of Edinburgh,
[email protected], www.dai.ed.ac.uk/homes/seeger/

Return to: SCS Technical Report Collection
School of Computer Science homepage

This page maintained by [email protected]