back to table of contents
   
title:
 
Towards an Axiomatization for the Generalization of the Kullback-Leibler Divergence to Belief Functions
publication:
 
EUSFLAT
part of series:
  Advances in Intelligent Systems Research
pages:   1090 - 1097
DOI:
  To be assigned soon (how to use a DOI)
author(s):
 
H¨¦l¨¨ne Soubaras
publication date:
 
July 2011
keywords:
 
Dempster-Shafer theory of belief functions, channel capacity, Kullback-Leibler divergence
abstract:
 
In his information theory, Shannon [1] defined a notion of uncertainty, the entropy, which has been generalized in several wways to belief functions [2]. He also defined the channel capacity for which we propose in this paper the first generalization to belief functions. To do that, we need first to generalize the Kullback-Leibler (KL) divergence, for which the present work proposes some axioms. Their list is still not exhaustive since the proposed solution is not unique. But there are many practical interests, since the notion of channel capacity is useful to characterize and optimize for example systems of sensors; its generalization to belief functions allows us to include imprecise sensors such as the human. Finally we show an example of gradient algorithm to compute the generalized channel capacity.
copyright:
 
© Atlantis Press. This article is distributed under the terms of the Creative Commons Attribution License, which permits non-commercial use, distribution and reproduction in any medium, provided the original work is properly cited.
full text: