BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:PAC Bayes - Alex Matthews\; Nikola Mrksic
DTSTART:20141127T150000Z
DTEND:20141127T163000Z
UID:TALK55678@talks.cam.ac.uk
CONTACT:39777
DESCRIPTION:The talk will be composed of two parts. The first part will be
  an introduction to PAC learning. The second will be an introduction to PA
 C-Bayes.\n\nProbably Approximately Correct learning (PAC learning) is a fr
 amework for the mathematical analysis of machine learning\, proposed in 19
 84 by Leslie Valiant. The framework introduces computational complexity th
 eory concepts to machine learning\, expecting the learner to find efficien
 t functions (polynomial time and space) using a polynomial learning proced
 ure as well. PAC learning gave rise to the field of computational learning
  theory\, whose primary goal is to compare the power of different learning
  models. This talk will introduce the basic concepts and present some of t
 he results obtained using PAC learning and VC theory. Since this part of t
 he talk is tutorial in nature no reading will be required.\n\nPAC-Bayes is
  a PAC like framework where the generalization error bounds are derived us
 ing a reference distribution chosen before seeing the data. The bounds are
  often very tight relative to other types of PAC bound. There are intimate
  connections to the Bayesian view of learning though the two theories are 
 not identical. In this talk we will give a tutorial on the basic concepts 
 of PAC-Bayes before discussing the now classic application to Gaussian pro
 cess classification. This part of the talk is tutorial in nature and relat
 ively self contained apart from some knowledge of Gaussian processes which
  will be assumed. An idea of the content of the talk can be gained by look
 ing at Seeger's work:\n\nhttp://www.jmlr.org/papers/volume3/seeger02a/seeg
 er02a.pdf\n\nbut a detailed understanding of this paper is certainly not e
 ssential to learn from the talk.
LOCATION:Engineering Department\, CBL Room 438
END:VEVENT
END:VCALENDAR