BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Drupal iCal API//EN
X-WR-CALNAME:Events items teaser
X-WR-TIMEZONE:America/Toronto
BEGIN:VTIMEZONE
TZID:America/Toronto
X-LIC-LOCATION:America/Toronto
BEGIN:DAYLIGHT
TZNAME:EDT
TZOFFSETFROM:-0500
TZOFFSETTO:-0400
DTSTART:20180311T070000
END:DAYLIGHT
BEGIN:STANDARD
TZNAME:EST
TZOFFSETFROM:-0400
TZOFFSETTO:-0500
DTSTART:20181104T060000
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
UID:69dba0e1b5db2
DTSTART;TZID=America/Toronto:20181121T121500
SEQUENCE:0
TRANSP:TRANSPARENT
DTEND;TZID=America/Toronto:20181121T121500
URL:https://uwaterloo.ca/data-systems-group/events/phd-seminar-distributed-
 dependency-discovery
LOCATION:DC - William G. Davis Computer Research Centre 200 University Aven
 ue West 1304 Waterloo ON N2L 3G1 Canada
SUMMARY:PhD Seminar • Distributed Dependency Discovery
CLASS:PUBLIC
DESCRIPTION:HEMANT SAXENA\, PHD CANDIDATE\nDavid R. Cheriton School of Comp
 uter Science\n\nWe address the problem of discovering dependencies from di
 stributed\nbig data.  Existing (non-distributed) algorithms focus on mini
 mizing\ncomputation by pruning the search space of possible dependencies.
  \nHowever\, distributed algorithms must also optimize data communication
 \ncosts\, especially in current shared-nothing settings.  To do this\, we
 \ndefine a set of primitives for dependency discovery\, which corresponds\
 nto data processing steps separated by communication barriers\, and we\npr
 esent efficient implementations that optimize both computation and\ncommun
 ication costs.  Using real data\, we show that algorithms built\nusing ou
 r primitives are significantly faster and more\ncommunication-efficient th
 an straightforward distributed\nimplementations.
DTSTAMP:20260412T134049Z
END:VEVENT
END:VCALENDAR