BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Generalized Multilevel Functional Regression - Ana-Maria Staicu\, 
 University of Bristol
DTSTART:20090609T133000Z
DTEND:20090609T143000Z
UID:TALK15499@talks.cam.ac.uk
CONTACT:Michael Sweeting
DESCRIPTION:We introduce Generalized Multilevel Functional Linear Models (
 GMFLM)\, a\nnovel statistical framework motivated by and applied to the Sl
 eep Heart\nHealth Study (SHHS)\, the largest community cohort study of sle
 ep. The\nprimary goal of SHHS is to study the association between sleep di
 srupted\nbreathing (SDB) and adverse health effects. An exposure of primar
 y\ninterest is the sleep electroencephalogram (EEG)\, which was observed f
 or\nthousands of individuals at two visits\, roughly 5 years apart. This u
 nique\nstudy design led to the development of models where the outcome\, e
 .g.\nhypertension\, is in an exponential family and the exposure\, e.g. sl
 eep\nEEG\, is multilevel functional data. We show that GMFLMs are\, in fac
 t\,\ngeneralized multilevel mixed effect models. Two consequences of this\
 nresult are that: 1) the mixed effects inferential machinery can be used\n
 for GMFLM and 2) functional regression models can be extended naturally to
 \ninclude\, for example\, additional covariates\, random effects and\nnonp
 arametric components. We propose and compare two inferential methods\nbase
 d on the parsimonious decomposition of the functional space.\n\nIn collabo
 ration with Ciprian M. Crainiceanu\, Chongzhi Di.
LOCATION:Large Seminar Room\, 1st Floor\, Institute of Public Health\, Uni
 versity Forvie Site\, Robinson Way\, Cambridge
END:VEVENT
END:VCALENDAR