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:Inference Group
SUMMARY:Latent Force Models with Gaussian Processes - Dr N
eil Lawrence (University of Manchester)
DTSTART;TZID=Europe/London:20100301T110000
DTEND;TZID=Europe/London:20100301T120000
UID:TALK23146AThttp://talks.cam.ac.uk
URL:http://talks.cam.ac.uk/talk/index/23146
DESCRIPTION:Physics based approaches to data modeling involve
constructing an accurate mechanistic model of data
\, often based on differential equations. Machine
learning approaches are typically data driven---p
erhaps through regularized function approximation.
\n\nThese two approaches to data modeling are ofte
n seen\nas polar opposites\, but in reality they a
re two different ends to a spectrum of approaches
we might take.\n\nIn this talk we introduce latent
force models. Latent force models are a new appro
ach to data representation that model data through
unknown forcing functions that drive differential
equation models. By treating the unknown forcing
functions with Gaussian process priors we can cre
ate probabilistic models that exhibit particular p
hysical characteristics of interest\, for example\
, in dynamical systems resonance and inertia. This
allows us to perform a synthesis of the data driv
en and physical modeling paradigms. We will show a
pplications of these models in systems biology and
modelling of human motion capture data.
LOCATION:TCM Seminar Room\, Cavendish Laboratory\, Departme
nt of Physics
CONTACT:Carl Scheffler
END:VEVENT
END:VCALENDAR