BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Memento EPFL//
BEGIN:VEVENT
SUMMARY:<b>A reward-modulated Hebbian learning rule can explain experiment
 ally observed network reorganization in a brain control task </b>
DTSTART:20091120T121500
DTSTAMP:20260407T210752Z
UID:3b73d0df7490a2b5fbfee4cbb5837a2ccaeaa6ed60506b1c86c4b1d3
CATEGORIES:Conferences - Seminars
DESCRIPTION:Robert Legenstein\, Institute for Theoretical Computer Science
 \, TU Graz\nIt has recently been shown in a brain-computer interface exper
 iment that motor cortical neurons change their tuning properties selective
 ly to compensate for errors induced by displaced decoding parameters. In p
 articular\, it was shown that the 3D tuning curves of neurons whose decodi
 ng parameters were re-assigned changed more than those of neurons whose de
 coding parameters had not been re-assigned. In this article\, we propose a
  simple learning rule that can reproduce this effect. Our learning rule us
 es Hebbian weight updates driven by a global reward signal and neuronal no
 ise. In contrast to most previously proposed learning rules\, this approac
 h does not require extrinsic information to separate noise from signal. Th
 e learning rule is able to optimize the performance of a model system with
 in biologically realistic periods of time under high noise levels. Further
 more\, when the model parameters are matched to data recorded during the b
 rain-computer interface learning experiments described above\, the model p
 roduces learning effects strikingly similar to those found in the experime
 nts. 
LOCATION:BC 01 https://plan.epfl.ch/?room==BC%2001
STATUS:CONFIRMED
END:VEVENT
END:VCALENDAR