Facebook pixel
Bocconi Home


Temi di ricerca

My research interests are essentially devoted to methodological aspects inherent to Bayesian statistical inference and to the comparison with the classical statistical inference, including fiducial inference.


In particular I am (or I have been) dealing with:

  • the meaning of a statistical model and its construction from a predictive point of view;
  • the characterization of default (non-informative) prior distributions from several points of view;
  •  improper priors and their relationships with finitely additive probability measure;
  • conjugate and enriched conjugate prior families for exponential models;
  • compatible priors for Bayesian models comparison (applications to the Hardy-Weinberg equilibrium model and to the linear models);
  • priors for testing simultaneously several types of hypothesis;
  • mixture priors for non-parametric Bayesian inference;
  • relationships between objective bayesian inference and fiducial inference.
Modificato il 03/09/2014