BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:&quot\;Exploring dependence between categorical variables: Benefit
 s and limitations of using variable selection within Bayesian clustering i
 n relation to searching for interactions&quot\; - Dr Michail Papathomas\, 
 University of St Andrews
DTSTART:20160607T133000Z
DTEND:20160607T143000Z
UID:TALK65256@talks.cam.ac.uk
CONTACT:Alison Quenault
DESCRIPTION:Detecting interactions when analyzing data sets created by lar
 ge cohort or association studies is becoming increasingly important in Bio
 statistics. Investigating complex dependence structures within a linear mo
 delling framework is not straightforward due to the difficulty in searchin
 g an unwieldy large space of competing models. One approach for reducing t
 he dimensionality of the problem is to utilize a Bayesian modelling approa
 ch based on the Dirichlet process. We investigate the relation between the
  Dirichlet process and linear modelling\, and discuss the utility of the D
 irichlet process for the exploration of high order interactions\, especial
 ly when sparse data are analysed.
LOCATION:Large  Seminar Room\, 1st Floor\, Institute of Public Health\, Un
 iversity Forvie Site\, Robinson Way\, Cambridge
END:VEVENT
END:VCALENDAR