Adapting synaptic weights (2)

As the previous video the purpose of this one is to show the possibility of self adapting weights for supervised training of a spiking neural network.
Here we have 100 presynaptic spiking neurons and hence more temporal resolution to codify the moment when the postsynaptic neuron must fire.
On the right you can see a raw plot of the synaptic values changing according to time: the more the master spiking neuron gets on the right the more the synapses tend to zero. Viceversa the synapses tend to grow when the master makes the spike toward the left.
The synaptic weights also tend to zero when the corresponding presynaptic neuron spikes after the master fired: very hebbian like.
Any comment would be appreciated, thankyou

Видео Adapting synaptic weights (2) канала MultiNeurons