Why Jordan algebras are natural in statistics: quadratic regression implies Wishart distributions - Université Toulouse III - Paul Sabatier - Toulouse INP Accéder directement au contenu
Article Dans Une Revue Bulletin de la société mathématique de France Année : 2011

Why Jordan algebras are natural in statistics: quadratic regression implies Wishart distributions

Jacek Wesolowski
  • Fonction : Auteur

Résumé

If the space Q of quadratic forms in R-n is splitted in a direct sum Q(1) circle plus (...) circle plus Q(k) and if X and Y are independent random variables of R-n, assume that there exist a real number a such that E(X vertical bar X Y) = a(X + Y) and real distinct numbers b(1), ..., b(k) such that E(q(X)vertical bar X + Y) = b(i)q(X + Y) for any q in Q(i). We prove that this happens only when k = 2, when R-n can be structured in a Euclidean Jordan algebra and when X and Y have Wishart distributions corresponding to this structure.

Dates et versions

hal-00666856 , version 1 (06-02-2012)

Identifiants

Citer

Gérard Letac, Jacek Wesolowski. Why Jordan algebras are natural in statistics: quadratic regression implies Wishart distributions. Bulletin de la société mathématique de France, 2011, 139 (1), pp.129-144. ⟨10.24033/bsmf.2603⟩. ⟨hal-00666856⟩
48 Consultations
0 Téléchargements

Altmetric

Partager

Gmail Facebook X LinkedIn More