PDF-2M.F.Balcan,Y.Liangandtherearenotheoreticalperformanceguaranteesforthe

DetectingHierarchicalCommunities3 Fig1IllustrationofancompactblobAnedgexymeansthatyisoneofxsnearestneighborsNotethatthenotionofcompactblobsisthesameastheclustersthatsatisfythegoodneighborh. This quest for better approximation algorithms is further fueled by the implicit hope that these better approximation also yield more accurate clusterings Eg for many prob lems such as clustering proteins by function or clustering images by subject The Mistake Bound model In this lecture we study the online learning protocol In this setting the following scenario is repeated inde57356nitely 1 The algorithm receives an unlabeled example 2 The algorithm predicts a classi57356cation of this examp cmuedu School of Computer Science Carnegie Mellon University Pittsburgh PA 152133891 Alina Beygelzimer beygelusibmcom IBM T J Watson Research Center Hawthorne NY 10532 John Langford jltticorg Toyota Technological Institute at Chicago Chicago IL 60637 gatechedu Department of Quantitative Economics Maastricht University heikoroeglinorg Computer Science Department University of Southern California shanghuatenggmailcom Abstract Motivated by the principle of agnostic learning we present an extension o Hubert Chan MohammadTaghi Hajiaghayi July 2007 CMUCS07142 School of Computer Science Carnegie Mellon University Pittsburgh PA 15213 School of Computer Science Carnegie Mellon University Pittsburgh PA ninamfavrimhuberthajiagha cscmuedu Abstract We co cmuedu School of Computer Science Carnegie Mellon University Pittsburgh PA 152133891 Alina Beygelzimer beygelusibmcom IBM T J Watson Research Center Hawthorne NY 10532 John Langford jltticorg Toyota Technological Institute at Chicago Chicago IL 60637 on . approximation. of . submodular. and XOS . functions by juntas. Vitaly. Feldman and Jan . Vondrâk. IBM . Research - . Almaden. Prelims:. Submodular. and XOS. Functions; . Learning. $20. $27. 1 MARIA - Carnegie Mellon University Pittsburgh, PA 15213 - 3891 ninamf@cs.cmu.edu www.cs.cmu.edu/~ninamf RESEARCH INTERESTS Learning Theory, Machine Learning , Theory of Computing, Artificial Intel Clustering. . Unsupervised Learning. Clustering, Informal Goals. Goal. : . Automatically . partition . unlabeled. . data into groups of similar . datapoints. .. . Question. : When and why would we want to do this?. Clustering. . Unsupervised Learning. Clustering, Informal Goals. Goal. : . Automatically . partition . unlabeled. . data into groups of similar . datapoints. .. . Question. : When and why would we want to do this?. of . submodular. and XOS . functions by juntas. Vitaly. Feldman and Jan . Vondrâk. IBM . Research - . Almaden. Prelims:. Submodular. and XOS. Functions; . Learning. $20. $27. Approximation by juntas. Clustering. . Unsupervised Learning. Clustering, Informal Goals. Goal. : . Automatically . partition . unlabeled. . data into groups of similar . datapoints. .. . Question. : When and why would we want to do this?. /11/2015. Active Learning . Supervised Learning. E.g., which . emails are spam and which are important.. E.g., . classify objects as chairs vs non . chairs.. Not . chair. chair. Not spam. spam. . Labeled Examples .