Journal of Convex Analysis 26 (2019), No. 3, 991--999
Copyright Heldermann Verlag 2019
A Note on the Approximate Symmetry of Bregman Distances
Stefan Kindermann
Industrial Mathematics Institute, Johannes Kepler University Linz, 4040 Linz, Austria
kindermann@indmath.uni-linz.ac.at
[Abstract-pdf]
\newcommand{\Breg}[1]{B_{#1}} The Bregman distance $\Breg{\xi_x}(y,x)$, $\xi_x \in \partial J(y),$ associated to a convex sub-differentiable functional $J$ is known to be in general non-symmetric in its arguments $x$, $y$. In this note we address the question when Bregman distances can be bounded against each other when the arguments are switched, i.e., if some constant $C>0$ exists such that for all $x,y$ on a convex set $M$ it holds that $\frac{1}{C} \Breg{\xi_x}(y,x) \leq \Breg{\xi_y}(x,y) \leq C \Breg{\xi_x}(y,x).$ We state sufficient conditions for such an inequality and prove in particular that it holds for the $p$-powers of the $\ell_p$ and $L^p$-norms when $1 < p <\infty$.
Keywords: Bregman distance, symmetry, strong monotonicity, convexity.
MSC: 52A41, 47H05, 26B25.
[ Fulltext-pdf (107 KB)] for subscribers only.