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:Path-space information metrics for uncertainty qua
ntification and coarse-graining of molecular syste
ms - Markos A. Katsoulakis (University of Massachu
setts)
DTSTART;TZID=Europe/London:20160614T150000
DTEND;TZID=Europe/London:20160614T160000
UID:TALK66443AThttp://talks.cam.ac.uk
URL:http://talks.cam.ac.uk/talk/index/66443
DESCRIPTION:We present path-space\, information theory-based\,
sensitivity analysis\, uncertainty quantification
and variational inference methods for complex hig
h-dimensional stochastic dynamics\, including chem
ical reaction networks with hundreds of parameters
\, Langevin-type equations and lattice kinetic Mon
te Carlo. We establish their connections with goal
-oriented methods in terms of new\, sharp\, uncert
ainty quantification inequalities that scale appro
priately at both long times and for high dimension
al state and parameter space. \; The combin
ation of proposed methodologies is capable to (a)
tackle non-equilibrium processes\, typically assoc
iated with coupled physicochemical mechanisms or b
oundary conditions\, such as reaction-diffusion pr
oblems\, and where even steady states are unknown
altogether\, e.g. do not have a Gibbs structure. T
he path-wise information theory tools\, \; (b)
yield a surprisingly simple\, tractable and easy-
to-implement approach to quantify and rank paramet
er sensitivities\, as well as \; (c) provide r
eliable parameterizations for coarse-grained molec
ular systems based on fine-scale data\, and ration
al model selection through path-space (dynamics-ba
sed) variational inference methods.
LOCATION:Seminar Room 2\, Newton Institute
CONTACT:INI IT
END:VEVENT
END:VCALENDAR