Activity Number:
|
455
|
Type:
|
Topic Contributed
|
Date/Time:
|
Wednesday, August 5, 2009 : 10:30 AM to 12:20 PM
|
Sponsor:
|
Section on Bayesian Statistical Science
|
Abstract - #304676 |
Title:
|
Population Monte Carlo and Cosmology
|
Author(s):
|
Darren Wraith*+
|
Companies:
|
Université de Paris Dauphine
|
Address:
|
CEREMADE, Paris cedex 16, Paris, International, 75775, France
|
Keywords:
|
importance sampling ; bayesian ; computational time ; cosmology
|
Abstract:
|
A significant appeal of importance sampling approaches, especially when compared to alternatives such as Markov Chain Monte Carlo (MCMC), lies in the possibility to use (massive) parallel sampling which considerably reduces the computational time involved in the estimation of parameters. In cosmology, such efficiency gains are often hugely advantageous as computational time using MCMC and related techniques can typically be very high or prohibitively so. In this presentation we apply the adaptive importance sampling technique of Population Monte Carlo as outlined in Cappé et al (2008) to simulated and actual data in cosmology and assess the potential computational time gains that can be made. Other benefits and issues associated with using such an approach for these problems, and more generally, are also explored.
|
- The address information is for the authors that have a + after their name.
- Authors who are presenting talks have a * after their name.
Back to the full JSM 2009 program |