BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:A Bayesian Composite Gaussian Process Model and its Application - 
 Thomas Santner (Ohio State University)
DTSTART:20180608T100000Z
DTEND:20180608T120000Z
UID:TALK107095@talks.cam.ac.uk
CONTACT:INI IT
DESCRIPTION:This talk will describe a flexible Bayesian model that can be 
 used to predict the output of a deterministic simulator code.  The model a
 ssumes that the output can be described as the sum of a smooth global tren
 d plus deviations from the global trend. The global trend and the local de
 viations are modeled as draws from independent GPs with separable correlat
 ion functions subject to appropriate constraints to enforce smoothness of 
 the global process compared with the local deviation process. The accuracy
  and limitations of predictions made using this model are demonstrated in 
 a series of examples.   The model is used to perform variable selection by
  identifying the most active inputs to the simulator. Inputs having ``smal
 ler&#39\;&#39\; posterior distributions of the model&#39\;s correlation pa
 rameters are judged to be more active.  A reference inactive input is adde
 d to the data to judge the size of the correlation parameter for inactive 
 inputs.    Joint work with Casey Davis and Christopher Hans
LOCATION:Seminar Room 2\, Newton Institute
END:VEVENT
END:VCALENDAR