BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Approximate nonlinear filtering with a neural network - Jean-Pasca
 l Pfister\, ETH Zurich
DTSTART:20151014T110000Z
DTEND:20151014T120000Z
UID:TALK61441@talks.cam.ac.uk
CONTACT:Guillaume Hennequin
DESCRIPTION:One of the most impressive property of the brain is its abilit
 y to continuously extract relevant features from the environment. For exam
 ple\, when surrounded by a noisy crowd\, humans can extract the voice of a
  single person or track its position over time. However\, it still remains
  unknown how this feature extraction takes place in the brain. We formulat
 e this problem in a very general framework where the task is to continuous
 ly infer hidden variables given past observations for a an arbitrary nonli
 near generative model. Even though\, the formal solution to this general p
 roblem has been given by the Kushner equation in the form of a stochastic 
 partial differential equation for the posterior distribution\, its practic
 al applicability remains difficult due to the closure problem (every momen
 t depends on higher order moments). Here\, we propose an approximate yet t
 ractable solution in the form of a sampling based filter. Interestingly\, 
 this sampling based filter is biologically plausible and could be implemen
 ted by a recurrent network of analog neurons. Furthermore\, we derive a le
 arning rule for the parameters of the model. We show through numerical sim
 ulations that the performance of this neural filter is as good as a standa
 rd particle filter in the limit of large number of particles. Remarkably\,
  when the number of dimensions is large and when the number of particles i
 s limited\, the neural filter outperforms the standard particle filter.
LOCATION:Cambridge University Engineering Department\, CBL\, BE-438 (http:
 //learning.eng.cam.ac.uk/Public/Directions)
END:VEVENT
END:VCALENDAR