BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:Representative Instances Selection
DTSTART:20160701T140000
DTEND:20160701T160000
DTSTAMP:20260406T185322Z
UID:7a3837ceda88932bf66ba7c039a689500256a3567a96ddbac46fc366
CATEGORIES:Conferences - Seminars
DESCRIPTION:Aleksei Triastcyn\nEDIC Candidacy Exam\nExam President: Prof. 
 Karl Aberer\nThesis Director: Prof. Boi Faltings\nCo-examiner: Prof. Matth
 ias Grossglauser\nBackground papers:Machine Teaching: An Inverse Problem t
 o Machine Learning and an Approach Toward Optimal Education.Submodularity 
 in Data Subset Selection and Active Learning.Leveraging for Big Data Regre
 ssion.Abstract\nWe are looking at the problem of finding the most represen
 tative examples in a dataset. Exponential growth of data nowadays makes\ni
 t difficult to process\, store\, or transfer information. Thus\, it would 
 be beneficial to select a small amount of representative examples\nthat ca
 n be used to train machine learning models without compromising the qualit
 y\, be transferred over network easily or stored for\nprolonged periods of
  time. In this proposal\, we consider three different perspectives: machin
 e teaching\, submodular data subset\nselection\, and leverage score sampli
 ng. We report the main ideas of each approach\, as well as their shortcomi
 ngs. Finally\, we outline\nthe future research possibilities\, potential s
 olutions to existing drawbacks\, and possible applications\, which include
  human teaching\,\nsecurity\, and multi-agent systems.
LOCATION:INR 212 http://plan.epfl.ch/?lang=en&room=INR212
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR