Bayesian Analysis

Universality of Bayesian Predictions

Alessio Sancetta

Full-text: Open access


This paper studies the theoretical properties of Bayesian predictions and shows that under minimal conditions we can derive finite sample bounds for the loss incurred using Bayesian predictions under the Kullback-Leibler divergence. In particular, the concept of universality of predictions is discussed and universality is established for Bayesian predictions in a variety of settings. These include predictions under almost arbitrary loss functions, model averaging, predictions in a non-stationary environment and under model misspecification.

Article information

Bayesian Anal., Volume 7, Number 1 (2012), 1-36.

First available in Project Euclid: 13 June 2012

Permanent link to this document

Digital Object Identifier

Mathematical Reviews number (MathSciNet)

Zentralblatt MATH identifier

Bayesian methods loss function model averaging structural change universal prediction


Sancetta, Alessio. Universality of Bayesian Predictions. Bayesian Anal. 7 (2012), no. 1, 1--36. doi:10.1214/12-BA701.

Export citation

See also

  • Related item: Bertrand Clarke. Comment on Article by Sancetta. Bayesian Anal., Vol. 7, Iss. 1 (2012), 37-44.
  • Related item: Feng Liang. Comment on Article by Sancetta. Bayesian Anal., Vol. 7, Iss. 1 (2012), 45-46.
  • Related item: Alessio Sancetta. Rejoinder. Bayesian Anal., Vol. 7, Iss. 1 (2012), 47-50.