Characterization of convolution splitting graphical models
Résumé
We aim at characterizing graphical models for convolution splitting distributions. Only marginal independence have been studied through the well-known Rao–Rubin condition. We generalize this condition for conditional independence and deduce the desired characterizations.
| Format | Figure, Image |
|---|---|
| Origine | Fichiers produits par l'(les) auteur(s) |