Wald Lecture: On the Bernstein-von Mises theorem with infinite-dimensional parameters
Freedman, David
Ann. Statist., Tome 27 (1999) no. 4, p. 1119-1141 / Harvested from Project Euclid
Text on Project Euclid
PDF
Alt PDF
If there are many independent, identically distributed observations governed by a smooth, finite-dimensional statistical model, the Bayes estimate and the maximum likelihood estimate will be close. Furthermore, the posterior distribution of the parameter vector around the posterior mean will be close to the distribution of the maximum likelihood estimate around truth. Thus, Bayesian confidence sets have good frequentist coverage properties, and conversely. However, even for the simplest infinite-dimensional models, such results do not hold. The object here is to give some examples.
Publié le : 1999-08-14
Classification:  Asymptotic confidence sets,  Bayesian inference,  consistency,  Gaussian priors,  62A15,  62C15 
@article{1017938917,
     author = {Freedman, David},
     title = {Wald Lecture: On the Bernstein-von Mises theorem with
			 infinite-dimensional parameters},
     journal = {Ann. Statist.},
     volume = {27},
     number = {4},
     year = {1999},
     pages = { 1119-1141},
     language = {en},
     url = {http://dml.mathdoc.fr/item/1017938917}
}

Freedman, David. Wald Lecture: On the Bernstein-von Mises theorem with
			 infinite-dimensional parameters. Ann. Statist., Tome 27 (1999) no. 4, pp.  1119-1141. http://gdmltest.u-ga.fr/item/1017938917/