Non-Informative Priors Via Sieves and Packing Numbers
Subhashis Ghosal, Jayanta K. Ghosh, Ravi Ramamoorthi · Birkhäuser Boston eBooks · 1997
In this paper, we propose methods for the construction of a non-informative prior through the uniform distributions on approximating sieves. In parametric families satisfying regularity conditions, it is shown that Jeffreys’ prior is obtained. The case with nuisance parameters is also considered. In the infinite dimensional situation, we show that such a prior leads to consistent posterior.