Seminars

back Back to all Seminars

AMCS Graduate Seminar | Bayesian Computational Statistics & Modeling: Some case studies and research topics

Start Date: February 2, 2017
End Date: February 2, 2017

By Professor Håvard Rue (KAUST)
 
In this overview talk, I will discuss the common theme of my new research group here at KAUST: Bayesian Computational Statistics & modelling. This theme is a mix of three topics, 'Bayes': which is the oldest take on statistics, 'computational statistics': how to design computational algorithms to make Bayesian inference about a class of models, and '(Bayesian) Modeling': how to think about and construct statistical models to solve a given problem. All these topics go hand in hand and interact with each other, and each one cannot be treated independently from the others. I will present some case studies which originates from our main project (www.r-inla.org), which should illustrate the generality and applicability of a specific class of models; Latent Gaussian Models. I will then discuss some ongoing research projects along these themes. I will conclude this talk presenting the new opportunities this new reseach group brings to KAUST, both with respect to courses offered and research topics (for MSc & PhD).
 
 
Biography: Professor Håvard Rue is a new faculty at KAUST (Jan 2017). He comes from the Norwegian University of Science and Technology, where he did his MSc in Marine Hydrodynamics (1988), PhD in Statistics (1993), and from 1994 a faculty member at the Department of Mathematical Sciences, and a Professor there since 1998. HeBB has been an associate editor for JRSS series-B, Scandinavian Journal of Statistics, Statistic Surveys, Annals of Statistics and Environmetrics, is is currently a co-editor of STAT. His research interest includes Bayesian computational statistics and modelling, which is summarised in R-INLA project, see www.r-inla.org. The main ingredient is Gaussian Markov random fields (GMRF) models, and with Leonhard Held, he wrote a monograph on the subject published by Chapman & Hall (2005). GMRFs is also the main ingredient doing (fast and accurate) approximate Bayesian analysis for latent Gaussian models using integrated nested Laplace approximations (INLA), which is published as a discussion paper for JRSS Series B in 2009 co-authored with S.Martino and N.Chopin. Recent results also put GMRFs into geostatistics using stochastic partial differential equations as the bridge, which provides an explicit link between certain Gaussian fields and GMRFs in triangulated lattices (published as a discussion paper for JRSS series B in 2011, with F.Lindgren and J.Lindstrøm). The Google Citation page, give more information about my research profile: see https://scholar.google.co.uk/citations?user=VJOn_ZkAAAAJ&hl=no
 

More Information:

For more info contact: Prof. Håvard Rue: email: haavard.rue@kaust.edu.sa
 
Date: Thursday 2nd Feb 2017
Time:12:00 PM - 01:00 PM
Location: Bldg. 9 - Lecture Hall 1
Refreshments: Lunch bags will be served