Characterization of convolution splitting graphical models

Jean Peyhardi (1, 2, 3) , Pierre Fernique (4, 1)

1 VIRTUAL PLANTS - Modeling plant morphogenesis at different scales, from genes to phenotype
2 IGF - Institut de Génomique Fonctionnelle
3 UM - Université de Montpellier
4 UMR AGAP - Amélioration génétique et adaptation des plantes méditerranéennes et tropicales

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.

Mots clés: Convolution splitting distribution; Rao-Rubin condition; graphical model

Domaines: Probabilités [math.PR]; Statistiques [math.ST]

Jean Peyhardi, Pierre Fernique. Characterization of convolution splitting graphical models. Statistics and Probability Letters, 2017, 126, ⟨10.1016/j.spl.2017.02.018⟩. ⟨hal-01286173⟩