On resolving the Savage–Dickey paradox - Archive ouverte HAL
Article Dans Une Revue Electronic Journal of Statistics Année : 2010

On resolving the Savage–Dickey paradox

Résumé

The Savage-Dickey ratio is known as a specialised representation of the Bayes factor (O'Hagan and Forster, 2004) that allows for a functional plugging approximation of this quantity. We demonstrate here that it is a generic approximation method instead of an identity imposing constraints on the prior distributions, while incidentally clarifying the measure-theoretic bases of the method. We provide furthermore a general framework to produce a converging approximation of the Bayes factor that is unrelated with the earlier approach of Verdinelli and Wasserman (1995).

Dates et versions

hal-00423846 , version 1 (13-10-2009)

Identifiants

Citer

Jean-Michel Marin, Christian P. Robert. On resolving the Savage–Dickey paradox. Electronic Journal of Statistics , 2010, 4, pp.643-654. ⟨10.1214/10-EJS564⟩. ⟨hal-00423846⟩
183 Consultations
0 Téléchargements

Altmetric

Partager

More