CMU-CS-01-109R
Computer Science Department
School of Computer Science, Carnegie Mellon University



CMU-CS-01-109R

Duality and Auxiliary Functions for Bregman Distances

Stephen Della Pietra, Vincent Della Pietra, John Lafferty

February 2002

This is a revised version of
Computer Science Technical Report CMU-CS-01-109,
published in October 2001.

CMU-CS-01-109R.ps
CMU-CS-01-109R.pdf


Keywords: Bragman distances, convex duality, Legendre functions, auxiliary functions


We formulate and prove a convex duality theorem for Bregman distances and present a technique based on auxiliary functions for deriving and proving convergence of iterative algorithms to minimize Bregman distance subject to linear constraints.

15 pages


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

This page maintained by reports@cs.cmu.edu