BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:The covariance perceptron: Theory and application to reservoir com
 puting - Matthieu Gilson\, Sofia Lawrie (Universitat Pompeu Fabra)
DTSTART:20201209T140000Z
DTEND:20201209T150000Z
UID:TALK154483@talks.cam.ac.uk
CONTACT:Yul Kang
DESCRIPTION:The dynamics in cortex is characterized by highly fluctuating 
 activity: Even under the very same experimental conditions the activity ty
 pically does not reproduce on the level of individual spikes. Given this v
 ariability\, how then does the brain realize its quasi-deterministic funct
 ion? This variability has first been seen as a limitation for the concept 
 of neural coding\, but recent advances have questioned this traditional vi
 ew and proposed possible functional roles for this variability. Here we pr
 opose a view where fluctuations are used as a basis to represent informati
 on. We show with the "covariance perceptron" how a fluctuation-based schem
 e can be powerful for the discrimination of spatio-temporal signals into s
 everal classes\, and compare it with the traditional mean-based classifica
 tion. Moreover\, we explore the application to reservoir computing\, where
  the information conveyed in the distinct statistical orders of time serie
 s is mixed\, and test its performance on the discrimination of spoken digi
 ts. Together\, this framework is a step toward biologically plausible lear
 ning rules for neuronal networks to extract complex statistical informatio
 n from sensory signals\, from principal component analysis to tensor facto
 rization.\n\nRef:\n\nGilson M\, Dahmen D\, Moreno-Bote R\, Insabato A\, He
 lias M (2020) The covariance perceptron: A new paradigm for classification
  and processing of time series in recurrent neuronal networks. PLoS Comput
  Biol 16(10): e1008127. "https://doi.org/10.1371/journal.pcbi.1008127":htt
 ps://doi.org/10.1371/journal.pcbi.1008127\n\nJoin Zoom Meeting\nhttps://us
 02web.zoom.us/j/84616631650?pwd=LzRNL1MwcFI1cXZkdTZqNUI5NTIrQT09\n\nMeetin
 g ID: 846 1663 1650\nPasscode: 072375
LOCATION:Online on Zoom
END:VEVENT
END:VCALENDAR