BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Differential geometry for representation learning - Georgios Arvan
 itidis
DTSTART:20211102T150000Z
DTEND:20211102T160000Z
UID:TALK164245@talks.cam.ac.uk
CONTACT:96082
DESCRIPTION:A common assumption in machine learning is that the data lie n
 ear a low dimensional manifold\, which the shortest paths between points s
 hould respect. In this talk\, we focus on differential geometry and presen
 t computational methods that enables us to learn and use this underlying s
 tructure. We rely on the latent space of generative models\, where we capt
 ure the geometry of the data manifold. We can then compute the associated 
 shortest paths\, which is a distance measure invariant under reparametriza
 tions of the latent space. We demonstrate though that this approach requir
 es to quantify meaningfully the uncertainty of the generative process. Fin
 ally\, we show that we can use the latent geometry in several ways\, as we
 ll as for applications in robotics and life sciences.
LOCATION:https://cl-cam-ac-uk.zoom.us/j/96995932839?pwd=RHVHN1d3Qy9XL2hxUl
 J2SGNPbTgxdz09
END:VEVENT
END:VCALENDAR