BEGIN:VCALENDAR
VERSION:2.0
PRODID:talks.ox.ac.uk
BEGIN:VEVENT
SUMMARY:Conditional Autoregressive models for disconnected graph and appli
cations - Anna Freni Sterrantino (Imperial College London)
DTSTART;VALUE=DATE-TIME:20190401T120000
DTEND;VALUE=DATE-TIME:20190401T130000
UID:https://talks.ox.ac.uk/talks/id/03fa803e-2394-4621-9e54-b31965434ee2/
DESCRIPTION:Bayesian Methods in disease mapping are widely used to estimat
e and smooth relative risk. Usually\, the underlying maps are defined usin
g an adjacency matrix based on the spatial neighbourhood of areal units\,
de facto defining connected graphs. In the presence of islands or disconti
nuous geographical regions\, disconnected graphs are created. Currently\,
these issues are solved by assigning the singletons to the closest areal u
nit and then the usual smoothing techniques are carried out to produce the
relative risk maps. This leads to incorrect relative risk estimates and r
esults in overfitting. But we define a scaled version for the intrinsic C
onditional autoregressive model on a map with islands and provide a clear
and unambiguous definition of the parameters and hyperparameters. To accou
nt for the islands\, we use a scaling option so the precision has the same
interpretation regardless of the particular structure of the map. This im
mediately suggests a fair prior for random effects associated with the isl
ands in a disconnected graph in terms of a normal distribution. These impr
ovements have been implemented in R-INLA. \nSpeakers:\nAnna Freni Sterrant
ino (Imperial College London)
LOCATION:Seminar Room 0
TZID:Europe/London
URL:https://talks.ox.ac.uk/talks/id/03fa803e-2394-4621-9e54-b31965434ee2/
BEGIN:VALARM
ACTION:display
DESCRIPTION:Talk:Conditional Autoregressive models for disconnected graph
and applications - Anna Freni Sterrantino (Imperial College London)
TRIGGER:-PT1H
END:VALARM
END:VEVENT
END:VCALENDAR