BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//UNIFR/WEBMASTER//NONSGML v1.0//EN
CALSCALE:GREGORIAN
BEGIN:VEVENT
DTSTART;VALUE=DATE:20231121T171500
DTEND;VALUE=DATE:20231121T171500
UID:14743@agenda.unifr.ch
DESCRIPTION:Today's developments in machine learning heavily focus on big data \napproaches. However, many applications in robotics require learning \napproaches that can rely on only few demonstrations or trials. The main \nchallenge boils down to finding structures that can be used in a wide \nrange of tasks, which requires us to advance on several fronts, \nincluding data structures and geometric structures.\n\nAs example of data structures, I will discuss the use of tensor \nfactorization techniques that can be used in global optimization \nproblems to efficiently extract and compress information, while \nproviding diverse human-guided learning capabilities (imitation and \nenvironment scaffolding). As examples of geometric structures, I will \ndiscuss the use of Riemannian geometry and geometric algebra in \nrobotics, where prior knowledge about the physical world can be embedded \nwithin the representations of skills and associated learning algorithms.\n
SUMMARY:Robot learning from few samples by exploiting the structure and geometry of data 
CATEGORIES:Colloque / Congrès / Forum
LOCATION:PER 08\, auditoire 2.52\, Chemin du Musée 3\, 1700 Fribourg
URL;VALUE=URI:https://agenda.unifr.ch/e/fr/14743
END:VEVENT
END:VCALENDAR