BEGIN:VCALENDAR
PRODID:-//eluceo/ical//2.0/EN
VERSION:2.0
CALSCALE:GREGORIAN
BEGIN:VEVENT
UID:www.tcs.tifr.res.in/event/34
DTSTAMP:20230914T125907Z
SUMMARY:Ranking Problems in Machine Learning: Theory and Applications
DESCRIPTION:Speaker: Shivani Agarwal\nComputer Science and Artificial Intel
ligence Laboratory\nMassachusetts Institute of Technology\n32 V\n\nAbstrac
t: \nIn the last few decades\, there has been considerable progress in the
understanding of binary classification (learning of binary-valued functio
ns) and regression (learning of real-valued functions)\, both classical pr
oblems in machine learning. Although several questions remain to be answer
ed\, there is a well-developed theory in place for these problems\, and pr
actical successes have been demonstrated in a variety of applications.\n\n
Recently\, a new class of learning problems\, namely ranking problems\, ha
ve begun to gain attention. In ranking\, one learns a real-valued function
that assigns scores to objects\, but the scores themselves do not matter
instead\, what is important is the relative ranking of objects induced by
those scores. Ranking problems arise in a variety of domains: in informat
ion retrieval\, one wants to rank documents according to relevance to some
topic or query in user-preference modeling\, one wants to rank items acc
ording to a user's likes and dislikes in computational biology\, one want
s to rank genes according to relevance to some disease. Ranking problems a
re mathematically distinct from both classification and regression\, and c
annot be analyzed using existing results for these problems.\n\nIn this ta
lk\, I will describe some recent results in both the theoretical understan
ding of ranking and its applications. In particular\, I will describe gene
ralization bounds for ranking algorithms based on the tools of uniform con
vergence and algorithmic stability\, and some preliminary results on the s
ample complexity of learning ranking functions. I will conclude with some
recent applications to ranking chemical structures for drug discovery.\n
URL:https://www.tcs.tifr.res.in/web/events/34
DTSTART;VALUE=DATE:20090930
LOCATION:A-212 (STCS Seminar Room)
END:VEVENT
END:VCALENDAR