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:Isaac Newton Institute Seminar Series
SUMMARY:Approximation of the Fisher information and design
in nonlinear mixed effects models - Mielke\, T (O
tto-von-Guericke)
DTSTART;TZID=Europe/London:20110812T110000
DTEND;TZID=Europe/London:20110812T114500
UID:TALK32331AThttp://talks.cam.ac.uk
URL:http://talks.cam.ac.uk/talk/index/32331
DESCRIPTION:The missing closed form representation of the prob
ability density of the observations is one main pr
oblem in the analysis of Nonlinear Mixed Effects M
odels. Often local approximations based on lineari
zations of the model are used to approximately des
cribe the properties of estimators. The Fisher Inf
ormation is of special interest for designing expe
riments\, as its inverse yields a lower bound of t
he variance of any unbiased estimator. Different l
inearization approaches for the model yield differ
ent approximations of the true underlying stochast
ical model and the Fisher Information (Mielke and
Schwabe (2010)). \nTarget of the presentation are
alternative motivations of Fisher-Information appr
oximations\, based on conditional moments. For an
individual design\, known inter-individual varianc
e and intra-individual variance\, the Fisher Infor
mation for estimating the population location para
meter vector results in an expression depending on
conditional moments\, such that approximations of
the expectation of the conditional variance and t
he variance of the conditional expectation yield a
pproximations of the Fisher Information\, which ar
e less based on distribution assumptions. \nTierne
y et. al. (1986) described fully exponential Lapla
ce approximations as an accurate method for approx
imating posterior moments and densities in Bayesia
n models. We present approximations of the Fisher
Information\, obtained by approximations of condit
ional moments with a similar heuristic and compare
the impact of different Fisher Information approx
imations on the optimal design for estimating the
population location parameters in pharmacokinetic
studies.\n\n
LOCATION:Seminar Room 1\, Newton Institute
CONTACT:Mustapha Amrani
END:VEVENT
END:VCALENDAR