|
Activity Number:
|
343
|
|
Type:
|
Contributed
|
|
Date/Time:
|
Tuesday, July 31, 2007 : 2:00 PM to 3:50 PM
|
|
Sponsor:
|
Section on Statistical Computing
|
| Abstract - #309823 |
|
Title:
|
Maximum Likelihood Estimation Using Iterative Importance Sampling
|
|
Author(s):
|
Fassil Nebebe*+ and Tak Mak
|
|
Companies:
|
Concordia University and Concordia University
|
|
Address:
|
, Montreal, QC, H3G1M8, Canada
|
|
Keywords:
|
Monte Carlo methods ; Latent variables ; importance functions ; approximating information matrix
|
|
Abstract:
|
Monte Carlo methods have been a popular alternative to high dimensional numerical integration in maximum likelihood inference involving incomplete data or latent variables. It is well known that the efficiency of Monte Carlo methods can be substantially increased by employing importance sampling. By using a class of importance functions with certain desirable properties, we consider the optimal choice of an importance function in a new setting of iterative importance sampling. We study the statistical properties of the resulting estimator using the present Monte Carlo optimization and examine also the issue of approximating the information matrix.
|
