BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Computational Neuroscience Journal Club - Matthew Kerr (CBL)
DTSTART:20161215T160000Z
DTEND:20161215T170000Z
UID:TALK69593@talks.cam.ac.uk
CONTACT:Daniel McNamee
DESCRIPTION:Matthew Kerr will cover:\n\n* Computational Precision of Menta
 l Inference as Critical Source of Human Choice Suboptimality\n* Jan Drugow
 itsch\, Valentin Wyart\, Anne-Dominique Devauchelle\, Etienne Koechlin\n* 
 Neuron (December 2016)\n* http://www.cell.com/neuron/fulltext/S0896-6273(1
 6)30843-1\n(abstract included below)\n\nAbstract:\nMaking decisions in unc
 ertain environments often requires combining multiple pieces of ambiguous 
 information from external cues. In such conditions\, human choices resembl
 e optimal Bayesian inference\, but typically show a large suboptimal varia
 bility whose origin remains poorly understood. In particular\, this choice
  suboptimality might arise from imperfections in mental inference rather t
 han in peripheral stages\, such as sensory processing and response selecti
 on. Here\, we dissociate these three sources of suboptimality in human cho
 ices based on combining multiple ambiguous cues. Using a novel quantitativ
 e approach for identifying the origin and structure of choice variability\
 , we show that imperfections in inference alone cause a dominant fraction 
 of suboptimal choices. Furthermore\, two-thirds of this suboptimality appe
 ar to derive from the limited precision of neural computations implementin
 g inference rather than from systematic deviations from Bayes-optimal infe
 rence. These findings set an upper bound on the accuracy and ultimate pred
 ictability of human choices in uncertain environments.
LOCATION:Cambridge University Engineering Department\, CBL\, BE-438 (http:
 //learning.eng.cam.ac.uk/Public/Directions)
END:VEVENT
END:VCALENDAR