BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Gaussian Processes at the Helm(holtz): A more fluid model for ocea
 n currents - Renato Berlinghieri\, Dept of Electrical Engineering and Comp
 uter Science\, MIT
DTSTART:20240116T160000Z
DTEND:20240116T170000Z
UID:TALK209695@talks.cam.ac.uk
CONTACT:Annabelle Scott
DESCRIPTION:Given sparse observations of buoy velocities\, oceanographers 
 are interested in reconstructing ocean currents away from the buoys and id
 entifying divergences in a current vector field. As a first and modular st
 ep\, we focus on the time-stationary case - for instance\, by restricting 
 to short time periods. Since we expect current velocity to be a continuous
  but highly non-linear function of spatial location\, Gaussian processes (
 GPs) offer an attractive model. But we show that applying a GP with a stan
 dard stationary kernel directly to buoy data can struggle at both current 
 reconstruction and divergence identification\, due to some physically unre
 alistic prior assumptions. To better reflect known physical properties of 
 currents\, we propose to instead put a standard stationary kernel on the d
 ivergence and curl-free components of a vector field obtained through a He
 lmholtz decomposition. We show that\, because this decomposition relates t
 o the original vector field just via mixed partial derivatives\, we can st
 ill perform inference given the original data with only a small constant m
 ultiple of additional computational expense. We illustrate the benefits of
  our method with theory and experiments on synthetic and real ocean data.\
 n\nPaper: https://proceedings.mlr.press/v202/berlinghieri23a/berlinghieri2
 3a.pdf.
LOCATION:Drum Building\, Madingley Rise Site\, West Cambridge and on zoom:
   https://zoom.us/j/6708259482?pwd=Qk03U3hxZWNJZUZpT2pVZnFtU2RRUT09
END:VEVENT
END:VCALENDAR