Posterior Sampling when the Normalizing Constant is Unknown

Stephen G. Walker (1)

1 IMSAS

This paper describes a means by which to undertake Bayesian posterior inference via sampling techniques when the normalizing constant is not computable and hence unavailable. The strategy relies on the introduction of latent variables which removes any integrals associated with the inaccessibility of the normalizing constant.

Mots clés: Physical Sciences

Domaines: Calcul [stat.CO]

Stephen G. Walker. Posterior Sampling when the Normalizing Constant is Unknown. Communications in Statistics - Simulation and Computation, 2011, 40 (05), pp.784-792. ⟨10.1080/03610918.2011.555042⟩. ⟨hal-00680015⟩