BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//talks.cam.ac.uk//v3//EN
BEGIN:VTIMEZONE
TZID:Europe/London
BEGIN:DAYLIGHT
TZOFFSETFROM:+0000
TZOFFSETTO:+0100
TZNAME:BST
DTSTART:19700329T010000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=-1SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:+0100
TZOFFSETTO:+0000
TZNAME:GMT
DTSTART:19701025T020000
RRULE:FREQ=YEARLY;BYMONTH=10;BYDAY=-1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
CATEGORIES:Statistics
SUMMARY:On the use of non-local priors for joint high-dime
nsional estimation and selection - David Rossel\,
University of Warwick
DTSTART;TZID=Europe/London:20150116T160000
DTEND;TZID=Europe/London:20150116T170000
UID:TALK56938AThttp://talks.cam.ac.uk
URL:http://talks.cam.ac.uk/talk/index/56938
DESCRIPTION:A main challenge in modern statistics is to devise
strategies that are effective in high dimensions.
Ideally one would like solutions which are on one
hand parsimonious\, i.e. help interpret the proce
ss that generated the data in an easy manner\, but
that at the same time yield accurate predictions.
As has been well-documented in the literature the
re is a tension between these two competing goals\
, e.g. simpler explanatory models tend to result i
n higher prediction errors\,\nand accurate predict
ive models tend to be more complex than one would
ideally wish for.\n\nWe explore the extent to whic
h these two goals can be reconciled\, adopting a B
ayesian framework and a novel formulation based on
non-local priors (NLPs).\nThis class of priors ha
s been proven to lead to faster learning rates for
model selection that are indispensable if one is
to attain Bayesian consistency in high-dimensions.
\nBecause they induce extra parsimony in the solut
ion\, NLPs typically result in adequately simple e
xplanatory models.\nInterestingly\, we recently di
scovered that NLPs also result in improved shrinka
ge rates for parameter estimation that lead to hig
hly accurate predictions in high dimensions\, henc
e providing models that are at the same time promi
sing for explanatory and predictive purposes.\nWe
will illustrate these issues by reviewing some of
the relevant theory and show various practical exa
mples\,\nas well as propose strategies to deal eff
iciently with some of the main computational issue
s at stake.
LOCATION:MR12\, Centre for Mathematical Sciences\, Wilberf
orce Road\, Cambridge
CONTACT:
END:VEVENT
END:VCALENDAR