Ricerca Eventi, seminari, short coursesShort CoursesBayesian Variable Selection: Historical Perspective and Recent Developments

Bayesian Variable Selection: Historical Perspective and Recent Developments

Marina Vannucci
Rice University

 

Mercoledì 15 dicembre 2021
10.00 - 12.00 - Lecture 1: Introduce Bayesian methods for variable selection that use spike-and-slab priors (discrete and continuous). Discuss structured priors and nonparametric constructions for applications in applied fields, such as high-throughput genomics and neuroimaging.
14.00 - 16.00 - Lecture 2: Cover extensions to non-Gaussian data and efficient sampling schemes for posterior inference. Show an application to non-homogeneous hidden Markov models. Conclude with a brief outlook on other aspects of variable selection priors, e.g., edge selection in graphical models.

 

Programma