Modelling spike-timing dependent plasticity.
Early on we had realized that the weight-change curve of ISO learing xxxcitexxx “looks similar” to the weight-change curves (often called “learning windows”) of spike-timing dependent plasticity (STDP, cite Markram, Bi&Pooxxx).
STPD is a form of correlation based (Hebbian) plasticity where a synaptic weight will grow if the pre-synaptic activity precedes the post-synaptic plasticity, while it will shrink if the order of these signals is turned around.
In this project we had developed a state-variable description of differential Hebbian learning (ISO-learning) and how this can be made comparable to STDP xxxcite Saudargienexxx.
The main focus of this study, however, was to show that synaptic plasticity can be a highly local process where the shape of the pre- and post-synaptic signals would influence the specific form of plasticity of a given synapse xxxcite Saudargienexxx.
This type of “Local Synaptic
Plasticity” (local learning) might take place at different
parts of the dendrite of a neuron and could be used to develop
different site-specific functional properties xxxcite Minijaxxx.
Main cooperation partners:
A. Saudargiene, M. Tamosiunaite, Kaunas
Belongs to Group(s):
Computational Neuroscience