BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:Approximate inference for  continuous time Markov processes
DTSTART:20120510T160000
DTEND:20120510T170000
DTSTAMP:20260507T130803Z
UID:912986fa2cefb8b5098d90d68f85343ec70ab8e9a8f350f34c0210be
CATEGORIES:Conferences - Seminars
DESCRIPTION:Prof. Manfred Opper from TU Berlin\nContinuous time Markov pro
 cesses (such as jump processes and diffusions) play an important role in t
 he modelling of dynamical systems in many\nscientific areas ranging from p
 hysics to systems biology.\n\nIn a variety of applications\, the stochasti
 c state of the system as a function of time is not directly observed. One 
 has only access to a set of nolsy observations\ntaken at discrete times. T
 he problem is then to infer the unknown state path as best as possible. In
  addition\, model parameters (like diffusion constants or transition\nrate
 s) may also be unknown and have to be estimated from the data. Since Monte
  Carlo sampling approaches can be time consuming one is interested in effi
 cient approximations. I will discuss variational approaches to this proble
 m which are based on methods developed in statistical physics and machine 
 learning and\nwhich have also interesting relations to stochastic optimal 
 control. Applications to transcriptional regulation in systems biology wil
 l be given.
LOCATION:INR 113 http://plan.epfl.ch/?zoom=20&recenter_y=5863814.94355&rec
 enter_x=730548.85489&layerNodes=fonds\,batiments\,labels\,information\,par
 kings_publics\,arrets_metro&floor=1&q=INR113
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR