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:National Centre for Statistical Ecology (NCSE) Sem
inars
SUMMARY:Inference for nonlinear dynamical systems\, with a
pplications to the ecology of infectious diseases.
- Ed Ionides\, University of Michigan\, USA.
DTSTART;TZID=Europe/London:20071114T160000
DTEND;TZID=Europe/London:20071114T170000
UID:TALK9187AThttp://talks.cam.ac.uk
URL:http://talks.cam.ac.uk/talk/index/9187
DESCRIPTION:Nonlinear stochastic dynamical models are used to
study ecological systems and many other systems oc
curing across the sciences and engineering. Such m
odels are natural to formulate and can be analyzed
mathematically and numerically. However\, difficu
lties associated with inference from time-series d
ata about unknown parameters in these models have
been a constraint on their application. We present
a new method that makes maximum likelihood estima
tion feasible for partially-observed nonlinear sto
chastic dynamical systems (also known as state-spa
ce models) where this was not previously the case.
The method is based on a sequence of filtering op
erations which are shown to converge to a maximum
likelihood parameter estimate. We make use of rece
nt advances in nonlinear filtering in the implemen
tation of the algorithm. We apply the method to th
e study of cholera in Bangladesh. We construct con
fidence intervals\, perform residual analysis\, an
d apply other diagnostics. Our analysis\, based up
on a model capturing the intrinsic nonlinear dynam
ics of the system\, reveals some effects overlooke
d by previous studies.
LOCATION:Center for Mathematical Sciences\, MR 4
CONTACT:Dr. Leah R Johnson
END:VEVENT
END:VCALENDAR