BEGIN:VCALENDAR
VERSION:2.0
PRODID:icalendar-ruby
CALSCALE:GREGORIAN
METHOD:PUBLISH
BEGIN:VTIMEZONE
TZID:Europe/Vienna
BEGIN:DAYLIGHT
DTSTART:20260329T030000
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=3
TZNAME:CEST
END:DAYLIGHT
BEGIN:STANDARD
DTSTART:20251026T020000
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
RRULE:FREQ=YEARLY;BYDAY=-1SU;BYMONTH=10
TZNAME:CET
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20260424T063011Z
UID:1763128800@ist.ac.at
DTSTART:20251114T150000
DTEND:20251114T160000
DESCRIPTION:Speaker: Jonathan Scott\nhosted by Veronika Sunko\nAbstract: Th
 e widespread use of powerful edge devices\, such as smartphones\, has led 
 to large scale decentralized data generation. Since this data is often sen
 sitive\, it cannot be centrally collected\, posing challenges to tradition
 al machine learning\, which relies on centralized datasets. Federated lear
 ning (FL) addresses this by training models locally on devices and only sh
 aring updates\, preserving privacy. However\, FL faces key challenges incl
 uding data and system heterogeneity\, high communication costs\, and limit
 ed device resources. This thesis presents a range of methods to improve fe
 derated learning\, with a primary focus on handling data heterogeneity und
 er realistic computational and communication constraints. In this talk we 
 present approaches that explicitly model and adapt to client diversity\, a
 s well as methods that personalize models to individual clients using hype
 rnetworks.
LOCATION:Central Bldg / O1 / Mondi 3 (I01.O1.010) and Zoom\, ISTA
ORGANIZER:
SUMMARY:Jonathan Scott: Thesis Defense: Data Heterogeneity and Personalizat
 ion in Federated Learning
URL:https://talks-calendar.ista.ac.at/events/6088
END:VEVENT
END:VCALENDAR