[Forum SIS] Bocconi DEC Seminar Sara Martino - October 15th at 1pm
Marco Bonetti
marco.bonetti a unibocconi.it
Mar 13 Ott 2009 12:54:39 CEST
Dear Colleagues,
The DEC Department of Bocconi University is pleased to invite you to
the seminar:
"Integrated Nested Laplace Approximations in practice"
held by
Sara Martino
Norwegian University of Science and Technology , Norway
Room 3-E4-SR03
Via Roentgen 1, Milan
Thursday, October 15th, 2009
1pm
=======
Abstract
Latent Gaussian models are a common construct in statistical
applications where
some latent field, which is assumed to have a Gaussian structure,
is undirectly observed through conditional independent
data.
This class of models is interesting as it can be sees an a framework
where many commonly used statistical models can be set into. The
latent Gaussian field can, in fact, be used to model, for instance, the
time and space dependence among data or the smooth effect of
covariates. Hence, smoothing-spline models, space time models,
semi-parametric regression, spatial and spatio-temporal models,
log-Gaussian Cox models, and geostatistical models, all fall into this
category.
Integrated Nested Laplace approximation (INLA) is a new approach to
Bayesian inference for latent Gaussian models when the main interest
is in
estimating posterior marginals.
INLA substitutes MCMC sampling with accurate deterministic
approximations of
such posterior marginals.
Its main advantage is speed: results are obtained in second and
minutes where
MCMC would require hours and days. Not less important is that INLA
allows for a
for a great deal of automation and parallelization of all
computational procedures.
In practice INLA can be used almost as a black-box to analyze latent
Gaussian models.
The program INLA, and its R interface make the INLA approach easily
available
for the end user and practically no programming at all is needed.
During this seminar we will illustrate the main ideas behind INLA and
illustrate, through
a series of examples, the range of models which, at the moment, can be
implemented
using the INLA program.
This is a joint work with Prof Haavard Rue (NTNU - Trondheim) and
Nicholas
Chopin (ENSAE - Paris)
======
You are all invited.
Sincerely,
Marco Bonetti
---
Marco Bonetti
Department of Decision Sciences
Bocconi University
Via Guglielmo Roentgen 1
20136 Milan, Italy
Tel +39 02 58365670
Fax +39 02 58365634
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://www.stat.unipg.it/pipermail/sis/attachments/20091013/9eeacc87/attachment-0001.html>
Maggiori informazioni sulla lista
Sis