On the Kullback-Leibler divergence between location-scale densities
F. Nielsen. (2019)cite arxiv:1904.10428Comment: 11 pages.
Abstract
We show that the $f$-divergence between any two densities of potentially
different location-scale families can be reduced to the calculation of the
$f$-divergence between one standard density with another location-scale
density. It follows that the $f$-divergence between two scale densities depends
only on the scale ratio. We then report conditions on the standard distribution
to get symmetric $f$-divergences. We illustrate this symmetric property with
the calculation of the Kullback-Leibler divergence between scale Cauchy
distributions. Finally, we show that the minimum $f$-divergence of any query
density of a location-scale family to another location-scale family is
independent of the query location-scale parameters.
Description
[1904.10428] On the Kullback-Leibler divergence between location-scale densities
%0 Journal Article
%1 nielsen2019kullbackleibler
%A Nielsen, Frank
%D 2019
%K divergences entropy information theory
%T On the Kullback-Leibler divergence between location-scale densities
%U http://arxiv.org/abs/1904.10428
%X We show that the $f$-divergence between any two densities of potentially
different location-scale families can be reduced to the calculation of the
$f$-divergence between one standard density with another location-scale
density. It follows that the $f$-divergence between two scale densities depends
only on the scale ratio. We then report conditions on the standard distribution
to get symmetric $f$-divergences. We illustrate this symmetric property with
the calculation of the Kullback-Leibler divergence between scale Cauchy
distributions. Finally, we show that the minimum $f$-divergence of any query
density of a location-scale family to another location-scale family is
independent of the query location-scale parameters.
@article{nielsen2019kullbackleibler,
abstract = {We show that the $f$-divergence between any two densities of potentially
different location-scale families can be reduced to the calculation of the
$f$-divergence between one standard density with another location-scale
density. It follows that the $f$-divergence between two scale densities depends
only on the scale ratio. We then report conditions on the standard distribution
to get symmetric $f$-divergences. We illustrate this symmetric property with
the calculation of the Kullback-Leibler divergence between scale Cauchy
distributions. Finally, we show that the minimum $f$-divergence of any query
density of a location-scale family to another location-scale family is
independent of the query location-scale parameters.},
added-at = {2019-12-11T14:19:29.000+0100},
author = {Nielsen, Frank},
biburl = {https://www.bibsonomy.org/bibtex/278fb84ae12045d10009cdd8bb323c23b/kirk86},
description = {[1904.10428] On the Kullback-Leibler divergence between location-scale densities},
interhash = {aec989f10002f940e09c77232ccceb3b},
intrahash = {78fb84ae12045d10009cdd8bb323c23b},
keywords = {divergences entropy information theory},
note = {cite arxiv:1904.10428Comment: 11 pages},
timestamp = {2019-12-11T14:19:29.000+0100},
title = {On the Kullback-Leibler divergence between location-scale densities},
url = {http://arxiv.org/abs/1904.10428},
year = 2019
}