BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:&quot\;How to sample a reliable neural estimate of the variable wo
 rld&quot\; - Mikko Juusola\, Professor of Systems Neuroscience\, Universit
 y of Sheffield
DTSTART:20161010T153000Z
DTEND:20161010T170000Z
UID:TALK67425@talks.cam.ac.uk
CONTACT:Lyn Dakin
DESCRIPTION:Abstract: The world is variable and dynamic. Its matter/energy
  is clustered into changing structures and events encompassing macro- and 
 micro-scales. The central problem facing all animals is how to best sample
  a reliable estimate of the world when the estimation itself is limited by
  variations in their neural machineries and by the uncertainty of their su
 rroundings. New results suggest that rather than working against variabili
 ty\, evolution works with it\, giving rise to reliable and robust informat
 ion sampling and representation in the nervous tissue. Photoreceptors samp
 le visual information stochastically and weight it against fluctuating res
 ponses of their neighbours. Such anti-aliased sampling improves neural est
 imates of intensity changes in image pixels. Visual interneurones further 
 adaptively sample and integrate synaptic information of photoreceptors to 
 improve their estimates of the structure of the world. I will present new 
 evidence for the hypothesis that variability in animals’ sensory systems
  is less noise and more a part of a solution to sample reliable estimates 
 of the variable world.\n
LOCATION:The Hodgkin Huxley Seminar Room\, Department of Physiology Develo
 pment and Neuroscience
END:VEVENT
END:VCALENDAR