BEGIN:VCALENDAR
VERSION:2.0
PRODID:ILLC Website
X-WR-TIMEZONE:Europe/Amsterdam
BEGIN:VTIMEZONE
TZID:Europe/Amsterdam
X-LIC-LOCATION:Europe/Amsterdam
BEGIN:DAYLIGHT
TZOFFSETFROM:+0100
TZOFFSETTO:+0200
TZNAME:CEST
DTSTART:19700329T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=-1SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:+0200
TZOFFSETTO:+0100
TZNAME:CET
DTSTART:19701025T030000
RRULE:FREQ=YEARLY;BYMONTH=10;BYDAY=-1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
UID:/NewsandEvents/Archives/2021/newsitem/12503/9-
 March-2021-Machine-learning-logic-and-structured-k
 nowledge-Balder-ten-Cate
DTSTAMP:20210302T162900
SUMMARY:Machine learning, logic, and structured kn
 owledge, Balder ten Cate
ATTENDEE;ROLE=Speaker:Balder ten Cate (Google Rese
 arch)
DTSTART;TZID=Europe/Amsterdam:20210309T181500
DTEND;TZID=Europe/Amsterdam:20210309T190000
LOCATION:Online via Zoom
DESCRIPTION:Over the last decade, advances in mach
 ine learning have taken the computer science commu
 nity by storm, enabling new applications and pushi
 ng the envelope on existing ones. Even on tasks th
 at are traditionally viewed as falling in the doma
 in of logical reasoning (e.g., reading comprehensi
 on tasks), deep neural models are now the state-of
 -the-art. Furthermore, logic and learning are perc
 eived by some as being distinct or even opposing a
 pproaches. At the same time, while various algorit
 hmic and hardware limitations that inhibited deep 
 learning solutions in the past have been successfu
 lly addressed, other fundamental problems arise, s
 uch as problems concerning fairness, explainabilit
 y, and controllability. In this talk, I will discu
 ss a few problems at the intersection of machine l
 earning and logic, including providing deep models
  with means to access structured knowledge.  Zoom 
 link: https://uva-live.zoom.us/j/82670894282
X-ALT-DESC;FMTTYPE=text/html:\n  <p>Over the last 
 decade, advances in machine learning have taken th
 e computer science community by storm, enabling ne
 w applications and pushing the envelope on existin
 g ones. Even on tasks that are traditionally viewe
 d as falling in the domain of logical reasoning (e
 .g., reading comprehension tasks), deep neural mod
 els are now the state-of-the-art. Furthermore, log
 ic and learning are perceived by some as being dis
 tinct or even opposing approaches. At the same tim
 e, while various algorithmic and hardware limitati
 ons that inhibited deep learning solutions in the 
 past have been successfully addressed, other funda
 mental problems arise, such as problems concerning
  fairness, explainability, and controllability. In
  this talk, I will discuss a few problems at the i
 ntersection of machine learning and logic, includi
 ng providing deep models with means to access stru
 ctured knowledge.</p>\n\n  <p>Zoom link: <a href="
 https://uva-live.zoom.us/j/82670894282" target="_b
 lank">https://uva-live.zoom.us/j/82670894282</a></
 p>\n
URL:https://research.google/people/107268/
END:VEVENT
END:VCALENDAR