|
GAUSSIAN MARKOV RANDOM FIELDS: AN INTRODUCTION:
***We still have some places left***
***Only 3 more weeks for early bird fees***
9:00 - 17:00, 29-30 September 2008
*) Course instructor:
Prof. Håvard Rue, Norwegian University of Science and Technology
*) Course Outline:
This course is aimed at PhD students and other academic staff who want
to understand and learn to apply and make use of Gaussian Markov
Random Fields (GMRFs) in Bayesian latent models.
The main motivation for GMRFs is its appliations to structured
additive regression models. These include (generalised) linear (mixed)
models, (generalised) additive
(mixed) models, smoothing-spline models, state-space models,
semiparametric regression models, spatial and spatio-temporal models,
log-Gaussian Cox-processes, geostatistical and geoadditive models. In
near all applications of these models, the latent field is
Gaussian. This enables us to treat all these models in the same
framework, which we name latent Gaussian models. In near all cases the
latent Gaussian is also a GMRF, so then we can take advantage of the
nice computational properties of GMRFs and derive fast(er) inference
algorithms using sparse matrix algorithms.
This allows for very quick Bayesian inference using integrated nested
Laplace
approximations. These deterministic approximations give ``practically
exact'' results compared to their MCMC alternatives, in just a small
fraction of the time.
This course will discuss GMRFs in general, how to do computation with
GMRFs using sparse matrix calculations, and MCMC and non-MCMC based
Bayesian inference for latent Gaussian models with examples. A tutorial
will
be given using the `inla' program and its R-interface.
*) Course requirements:
Some background in Bayesian statistics, modelling and computational
based inference.
*) Location:
The course will take place in the MSc room of the Division of
Epidemiology, Public Health and Primary Care of the Faculty of
Medicine and the participants will be able to use these facilities. Note
that this is in the Faculty of Medicine, Imperial College London at St.
Mary's Campus, and NOT in the main campus in South Kensington.
The complete address is
Faculty of Medicine
Imperial College London
St. Mary's Campus, Norfolk Place
W2 1PG London - UK
More information on how to arrive can be found at
http://www1.imperial.ac.uk/medicine/contacts/campuses/stmarys/
*) Booking and course fees:
Early bird- by July 31st 2008
* £90 for postgraduate students registered at academic institutions
* £140 for Imperial College staff
* £180 for all other participants
Late registration - After July 31st 2008
* £100 for postgraduate students registered at academic institutions
* £160 for Imperial College staff
* £220 for all other participants
This includes the course materials, refreshment and lunch.
*) In order to book a place for the course, you should contact Sara
Geneletti ([log in to unmask]).
Summary of course contents
Day 1
* Registration (9:00-9:30)
* Morning Session (9:30-12:30)
Background on GMRFs
* Coffee break (10:30-10:45)
Sparse matrix calculations
Latent GMRF models
* Lunch (12:30 - 14:00)
* Afternoon Session (14:00-17:00)
MCMC for latent GMRF models with examples
* Coffee Break (15:45-16:00)
Non-MCMC Bayesian inference for latent GMRF models
Day 2
* Morning Session (9:30-12:30)
The `inla' program with examples
* Coffee break (10:30-10:45)
The R-inla interface to inla, with examples
* Lunch (12:30 - 14:00)
* Afternoon Session (14:00-17:00)
Discussion of participants problems
*) More Info
You can find updated information about the course at the following web
site:
http://www.bias-project.org.uk/GMRFCourse/
GAUSSIAN MARKOV RANDOM FIELDS: AN INTRODUCTION:
***Only 3 more weeks for early bird fees***
9:00 - 17:00, 29-30 September 2008
*) Course instructor:
Prof. Håvard Rue, Norwegian University of Science and Technology
*) Course Outline:
This course is aimed at PhD students and other academic staff who want
to understand and learn to apply and make use of Gaussian Markov
Random Fields (GMRFs) in Bayesian latent models.
The main motivation for GMRFs is its appliations to structured
additive regression models. These include (generalised) linear (mixed)
models, (generalised) additive
(mixed) models, smoothing-spline models, state-space models,
semiparametric regression models, spatial and spatio-temporal models,
log-Gaussian Cox-processes, geostatistical and geoadditive models. In
near all applications of these models, the latent field is
Gaussian. This enables us to treat all these models in the same
framework, which we name latent Gaussian models. In near all cases the
latent Gaussian is also a GMRF, so then we can take advantage of the
nice computational properties of GMRFs and derive fast(er) inference
algorithms using sparse matrix algorithms.
This allows for very quick Bayesian inference using integrated nested
Laplace
approximations. These deterministic approximations give ``practically
exact'' results compared to their MCMC alternatives, in just a small
fraction of the time.
This course will discuss GMRFs in general, how to do computation with
GMRFs using sparse matrix calculations, and MCMC and non-MCMC based
Bayesian inference for latent Gaussian models with examples. A tutorial
will
be given using the `inla' program and its R-interface.
*) Course requirements:
Some background in Bayesian statistics, modelling and computational
based inference.
*) Location:
The course will take place in the MSc room of the Division of
Epidemiology, Public Health and Primary Care of the Faculty of
Medicine and the participants will be able to use these facilities. Note
that this is in the Faculty of Medicine, Imperial College London at St.
Mary's Campus, and NOT in the main campus in South Kensington.
The complete address is
Faculty of Medicine
Imperial College London
St. Mary's Campus, Norfolk Place
W2 1PG London - UK
More information on how to arrive can be found at
http://www1.imperial.ac.uk/medicine/contacts/campuses/stmarys/
*) Booking and course fees:
Early bird- by July 31st 2008
* £90 for postgraduate students registered at academic institutions
* £140 for Imperial College staff
* £180 for all other participants
Late registration - After July 31st 2008
* £100 for postgraduate students registered at academic institutions
* £160 for Imperial College staff
* £220 for all other participants
This includes the course materials, refreshment and lunch.
*) In order to book a place for the course, you should contact Sara
Geneletti ([log in to unmask]).
Summary of course contents
Day 1
* Registration (9:00-9:30)
* Morning Session (9:30-12:30)
Background on GMRFs
* Coffee break (10:30-10:45)
Sparse matrix calculations
Latent GMRF models
* Lunch (12:30 - 14:00)
* Afternoon Session (14:00-17:00)
MCMC for latent GMRF models with examples
* Coffee Break (15:45-16:00)
Non-MCMC Bayesian inference for latent GMRF models
Day 2
* Morning Session (9:30-12:30)
The `inla' program with examples
* Coffee break (10:30-10:45)
The R-inla interface to inla, with examples
* Lunch (12:30 - 14:00)
* Afternoon Session (14:00-17:00)
Discussion of participants problems
*) More Info
You can find updated information about the course at the following web
site:
http://www.bias-project.org.uk/GMRFCourse/
|
