BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:Learning Data Representations with Convex Optimization
DTSTART:20090406T161500
DTSTAMP:20260407T130045Z
UID:999a54dd7de7372a9c497ededdb653eb1165ce1a983ecb2e9c24b305
CATEGORIES:Conferences - Seminars
DESCRIPTION:Dr. Kilian Weinberger\, Yahoo\, Mountain View\, USA\nOne of th
 e fundamental challenges of machine learning and artificial intelligence i
 s the learning of suitable representations of data. Many learning algorith
 ms assume that input data is presented in low-dimensional vectorial form\,
  where Euclidean distances reflect dissimilarities. However\, different ta
 sks can have very different definitions of similarity. Ideally\, one shoul
 d use a "hand-tailored" representation of each particular data set for any
  given task. \n\nIn this talk\, I will give an overview of three algorithm
 s for learning compact representations that give rise to semantically mean
 ingful similarity metrics. Each of the algorithms involves\, at its core\,
  a convex optimization problem that learns the new representation under me
 aningful constraints. This framework provides perfect reproducibility and 
 theoretical guarantees.  \n\nThe three methods are suited to different dat
 a settings: Taxonomy Embedding learns low-dimensional representations of t
 ext documents in a Euclidean space such that distances reflect dissimilari
 ties according to a given topic hierarchy. Maximum Variance Unfolding redu
 ces the dimensionality of data sets with underlying manifold structure and
  can be viewed as a non-linear extension of Principal Component Analysis (
 PCA). Large Margin Nearest Neighbor learns a relaxation of the Euclidean m
 etric tailored specifically for k-nearest neighbor classification. \n\nI p
 resent state-of-the-art classification results on several real world appli
 cations\, including face recognition and document categorization on the OH
 SUMED medical journal data base.\n\n\nBio:\n\nKilian Weinberger is a Resea
 rch Scientist at Yahoo Research in Santa Clara\, California. He works on n
 ext-generation spam filtering algorithms\, multimedia search\, high-dimens
 ional data analysis\, and machine learning with convex optimization. He re
 ceived his Ph.D. in Computer Science at the University of Pennsylvania in 
 2007 under the supervision of Professor Lawrence Saul. His work on supervi
 sed and unsupervised metric learning has won several outstanding paper awa
 rds at CVPR\, AISTATS and ICML. Prior to his doctoral studies he earned a 
 first class honor BA in Mathematics and Computer Science from the Universi
 ty of Oxford. Kilian is originally from Germany. \n\nK. Weinberger's homep
 age
LOCATION:INM202
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR