BEGIN:VCALENDAR
VERSION:2.0
PRODID:icalendar-ruby
CALSCALE:GREGORIAN
X-WR-CALNAME:ECE MS Thesis Defense: &quot\;Distributing Frank-Wolfe via Map
 -Reduce\,&quot\; Armin Moharrer
X-WR-TIMEZONE:Eastern Time (US & Canada)
BEGIN:VEVENT
DTSTAMP:20260916T152055Z
UID:tag:localist.com\,2008:EventInstance_3488378
DTSTART:20180418T133000Z
DTEND:20180418T143000Z
DESCRIPTION:Abstract: Large-scale optimization problems abound in data mini
 ng and machine learning applications\, and the computational challenges th
 ey pose are often addressed through parallelization. We identify structura
 l properties under which a convex optimization problem can be massively pa
 rallelized via map-reduce operations using the Frank-Wolfe (FW) algorithm.
  The class of problems that can be tackled this way is quite broad and inc
 ludes experimental design\, AdaBoost\, and projection to a convex hull. Im
 plementing FW via map-reduce eases parallelization and deployment via comm
 ercial distributed...
LOCATION:
SUMMARY:ECE MS Thesis Defense: &quot\;Distributing Frank-Wolfe via Map-Redu
 ce\,&quot\; Armin Moharrer
URL;VALUE=URI:https://calendar.northeastern.edu/event/ece_ms_thesis_defense
 _quotdistributing_frank-wolfe_via_map-reducequot_armin_moharrer
END:VEVENT
END:VCALENDAR