BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Bayesian regression models for complex spatially or serially corre
 lated functional data - Jeffrey Morris\, MD Anderson Cancer Center
DTSTART:20181026T150000Z
DTEND:20181026T160000Z
UID:TALK109654@talks.cam.ac.uk
CONTACT:Dr Sergio Bacallado
DESCRIPTION:A series of Bayesian regression modeling strategies that can b
 e used for spatially or longitudinally correlated functional data will be 
 described. The methods are intended for use with complex functional data\,
  measured on fine grids and with complexities including multi-dimensional 
 and possibly non-Euclidean domains\, local features like changepoints and 
 peaks\, and sampled on high-dimensional fine grids.  Intrafunctional corre
 lation is handled through basis function modeling\, while interfunctional 
 correlation is captured by one of three approaches: (1) parametric or nonp
 arametric random effect functions\, (2) separable or non-separable spatial
  (or temporal) inter-functional processes\, or (3) tensor-basis function m
 odeling. Rigorous Bayesian inference is done in such a way that adjusts fo
 r any potential multiple testing issues. We will describe these general ap
 proaches and illustrate them on a series of complex\, high-dimensional\, s
 patially and longitudinally correlated functional data sets coming from st
 rain tensor data from a glaucoma study\, bladder cancer genomic maps\, eve
 nt-related potential data from a smoking cessation study. We will also dis
 cuss recent work in which we have developed spatiotemporal quantile functi
 onal regression approaches that we are applying to model temporal climate 
 change in terms of intraseasonal temperature and precipitation distributio
 ns.  Full Bayesian infererence is available for answering inferential ques
 tions while accounting for multiple testing according to experimentwise er
 ror rate and/or false discovery rate criteria.  These methods are encapsul
 ated within the BayesFMM package of general Bayesian functional mixed mode
 ls\, for which we are developing general software in R and Matlab.
LOCATION:MR12
END:VEVENT
END:VCALENDAR