|
Activity Number:
|
399
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Wednesday, August 6, 2008 : 10:30 AM to 12:20 PM
|
|
Sponsor:
|
Section on Bayesian Statistical Science
|
| Abstract - #301521 |
|
Title:
|
New Perspectives on the Estimation of Normalizing Constants via Posterior Simulations
|
|
Author(s):
|
Giovanni Petris*+ and Luca Tardella
|
|
Companies:
|
University of Arkansas and Università di Roma "La Sapienza"
|
|
Address:
|
SCEN 301 Mathematical Sciences, Fayetteville, AR, 72701,
|
|
Keywords:
|
Normalizing constant ; Integrated likelihood ; MCMC
|
|
Abstract:
|
In this paper we recast the problem of evaluating the normalizing constant of an arbitrary density function with the aid of an arbitrary MC or MCMC sampling scheme. One new basic idea is that of perturbing the original target density function, whose normalizing constant has to be evaluated, in such a way that the perturbed density has the same original normalizing constant plus a known arbitrary positive mass. We focus on the effectiveness of the estimators derived from this new idea when the perturbed density is obtained via the Hyperplane Inflation idea of Petris and Tardella. The proposed estimators share the original simplicity of the harmonic mean estimator, yielding consistent MC or MCMC estimators based only on a simulated sample from the distribution proportional to the original density. The calibration of the optimal choice of the arbitrary positive mass is also discussed.
|
