BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Hybrids of Generative and Discriminative Models (Cont) - Simon Lac
 oste-Julien (University of Cambridge)
DTSTART:20090226T140000Z
DTEND:20090226T153000Z
UID:TALK15409@talks.cam.ac.uk
CONTACT:Shakir Mohamed
DESCRIPTION:In this RCC\, I will cover the topic of combining generative a
 nd discriminative learning approaches for classification.\nI will continue
  where we left off two weeks ago on the topic of combining generative and 
 discriminative learning. We will look at the Bishop's paper and I will bri
 efly present the Bouchard's paper\, contrasting the frequentist philosophy
  with the Bayesian one. For the Bouchard's paper\, only look at the first 
 two sections and maybe quickly browse over the experimental section - the 
 theoretical section is full of typos and most probably not correct.\n\nBel
 ow is the summary from two weeks ago for your reference:\n\n-----------\n\
 nI will start by covering a classic:\n* _On Discriminative vs. Generative 
 Classifiers: A comparison of logistic regression and Naive Bayes_\, Andrew
  Y. Ng and Michael Jordan.\nIn NIPS 14\, 2002.\n"pdf":http://ai.stanford.e
 du/~ang/papers/nips01-discriminativegenerative.pdf\n\nThis paper is unfort
 unately not an easy read\, but it is one of the most cited on the topic of
  comparing generative and discriminative learning. The analysis is somewha
 t specific\, but still provides interesting insight.\n\nI will then compar
 e the two hybrid generative / discriminative approaches proposed in the fo
 llowing two papers (the first one being more Bayesian\; the second one bei
 ng more Frequentist):\n* _Generative or Discriminative? getting the best o
 f both worlds_ \, Bishop\, C. M. and Lasserre\, J. (2007) In Bayesian Stat
 istics 8\, Bernardo\, J. M. et al. (Eds)\, Oxford University Press. 3–23
 . With discussion. "pdf":http://research.microsoft.com/en-us/um/people/cmb
 ishop/downloads/Bishop-Valencia-07.pdf\n* _Bias-variance tradeoff in hybri
 d generative-discriminative models_\, G. Bouchard.  In proc. of the Sixth 
 International conference on Machine Learning and Applications (ICMLA 07)\,
  Cincinnati\, Ohio\, USA\,\n13-15 Dec.2007.\, 2007.\n"pdf":http://eprints.
 pascal-network.org/archive/00003070/01/bouchard-GenerativeDiscrim.pdf\n\nI
 f you have time to read only one paper\, read the Bishop's one (I will foc
 us more on it) and skim through the other two. A slightly shorter conferen
 ce version is available "here":http://mi.eng.cam.ac.uk/~jal62/publis/cvpr0
 6_hybrid_models.pdf\n.\n\nIf you are quite interested on the topic and you
  want more details\, have a look at the "PhD thesis of Julia Lassere":http
 ://mi.eng.cam.ac.uk/~jal62/publis/thesis.pdf and chapter\n5 of the "PhD th
 esis of Guillaume\nBouchard":http://www.math.u-psud.fr/select/pub/thesegb.
 pdf (chapter 5 is in English even though the several other chapters are in
  French).\n\n
LOCATION:Engineering Department\, CBL Room 438
END:VEVENT
END:VCALENDAR