BEGIN:VCALENDAR
VERSION:2.0
PRODID:icalendar-ruby
CALSCALE:GREGORIAN
METHOD:PUBLISH
BEGIN:VTIMEZONE
TZID:Europe/Vienna
BEGIN:DAYLIGHT
DTSTART:20180325T030000
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=3
TZNAME:CEST
END:DAYLIGHT
BEGIN:STANDARD
DTSTART:20181028T020000
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=10
TZNAME:CET
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20180322T194417Z
UID:5a99531bcaf97320968383@ist.ac.at
DTSTART:20180405T090000
DTEND:20180405T100000
DESCRIPTION:Speaker: Shahar Mendelson\nhosted by Jan Maas\nAbstract: Proble
ms in modern statistics involve large data sets\, and analyzing such sets
is one of the main challenges of statistical learning theory. Over the las
t 15 years it has become clear that such problems have strong connections
with high dimensional geometry\, and the study of those connections has le
d to significant progress in our understanding of statistical learning the
ory.I will present a few of the ideas behind this geometric viewpoint to l
earning\, and as an example\, I will describe the instrumental role high d
imensional geometry played in the solution of one of the classical questio
ns in statistics - the prediction problem - with the introduction of a le
arning procedure that performs with the optimal sample complexity for (alm
ost) any prediction problem.
LOCATION:Mondi Seminar Room 2\, Central Building\, IST Austria
ORGANIZER:pdelreal@ist.ac.at
SUMMARY:Statistical learning theory from a geometric viewpoint
URL:https://talks-calendar.app.ist.ac.at/events/1175
END:VEVENT
END:VCALENDAR