BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Graphical Gaussian Models with Edge and Vertex Symmetries - Steffe
 n Lauritzen University of Oxford
DTSTART:20090529T150000Z
DTEND:20090529T160000Z
UID:TALK18119@talks.cam.ac.uk
CONTACT:8047
DESCRIPTION:Graphical Gaussian Models are typically defined through restri
 cting\nelements of the inverse covariance matrix\, aka the concentration\n
 matrix\, to be equal to zero\; this corresponds to conditional\nindependen
 ce restrictions of an undirected graph determined by absence\nof edges bet
 ween pairs of variables with zero joint concentration. In\nthis way\, comp
 lex multivariate distributions may be described with\nparsimony. This lect
 ure is concerned with further restrictions on the\ndistribution determined
  by symmetry\; the symmetry restrictions are\nobtained by partitioning the
  vertex set into vertex colour classes and\nthe edge set into edge colour 
 classes so that parameters corresponding\nto objects in the same colour cl
 ass are restricted to be identical.\nThe lecture will describe examples of
  such models and discuss their\nproperties. The material presented is base
 d on joint work with Sren\nHjsgaard and Helene Neufeld.\n
LOCATION:MR12\, CMS\, Wilberforce Road\, Cambridge\, CB3 0WB
END:VEVENT
END:VCALENDAR