 |
 |
|
|
| Detailed view | Preprint, Working Paper, ... |
 |
|
|
| Available versions: | v1 (2006-04-11) | v2 (2006-11-13) |
|
|
| Attached file list to this document: | ||||||||||
|
||||||||||
|
|
|
| Normalized information-based divergences |
|
|
| Jean-François Coeurjolly1Rémy Drouilhet1Jean-François Robineau1 |
|
|
|
| This paper is devoted to the mathematical study of some divergences based on the mutual information well-suited to categorical random vectors. These divergences are generalizations of the ``entropy distance" and ``information distance". Their main characteristic is that they combine a complexity term and the mutual information. We then introduce the notion of (normalized) information-based divergence, propose several examples and discuss their mathematical properties in particular in some prediction framework. |
|
|
|
|
|
|
|
|
|
| 1: | LABSAD - Laboratoire de Statistiques et Analyse des Données |
|
|
|
|
|
|
|
| Information theory – entropy distance – information distance – triangular inequality |
| hal-00022566, version 2 | |
| http://hal.archives-ouvertes.fr/hal-00022566 | |
| oai:hal.archives-ouvertes.fr:hal-00022566 | |
| From: Jean-François Coeurjolly | |
| Submitted on: Monday, 13 November 2006 09:24:12 | |
| Updated on: Monday, 13 November 2006 09:48:55 | |
