BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Computational Neuroscience Journal Club - Professor Daniel Wolpert
  ( Department of Engineering)
DTSTART:20100504T150000Z
DTEND:20100504T160000Z
UID:TALK24761@talks.cam.ac.uk
CONTACT:Dr Jean-Pascal Pfister
DESCRIPTION:Daniel Wolpert will  present\n\nHuman Active Learning\nR. Cast
 ro\, C. Kalish\, R. Nowak\, R. Qian\, T. Rogers and X. Zhu\,\nAdvances in 
 Neural Information Processing Systems (NIPS) 2008\n\nThe paper is availabl
 e on  http://dl.dropbox.com/u/4193800/Castro_NIPS_2008.pdf\n\n\nAbstract:\
 nWe investigate a topic at the interface of machine learning and cognitive
  science. Human active learning\, where learners can actively query the wo
 rld for information\, is contrasted with passive learning from random exam
 ples. Furthermore\, we compare human active learning performance with pred
 ictions from statistical learning theory. We conduct a series of human cat
 egory learning experiments inspired by a machine learning task for which a
 ctive and passive learning error bounds are well understood\, and dramatic
 ally distinct. Our results indicate that humans are capable of actively se
 lecting informative queries\, and in doing so learn better and faster than
  if they are given random training data\, as predicted by learning theory.
  However\, the improvement over passive learning is not as dramatic as tha
 t achieved by machine active learning algorithms. To the best of our knowl
 edge\, this is the first quantitative study comparing human category learn
 ing in active versus passive settings.\n
LOCATION:Cambridge University Engineering Department\, CBL Rm #438 (http:/
 /learning.eng.cam.ac.uk/Public/Directions)
END:VEVENT
END:VCALENDAR