Introduction on Spiking Neural Networks (SNNs) by using PyNN on the SpiNNaker neuromorphic system.
- make the EBRAINS credentials to access the SpiNNaker server (https://spinn-20.cs.man.ac.uk/hub/login)
- login on the Jupyter Lab interface
- clone this repository
git clone https://github.com/albertoarturovergani/CNT-2021 - Open the directory
SpiNNaker/and run the CNT notebook
https://univ-amu-fr.zoom.us/j/94280459110?pwd=b25zQytlQ1dIK0x2OTU5OXQ3dzFEZz09
contact: alberto.vergani@univ-amu.fr
- neurons
- cell types
- populations
- recording variables
- connections
- synapse types
- connections types
- projections
- simulation managing
- computational settings
- save and load outputs
- visualization tools
- 1D entry network
- 1D decaying network
- 1D persistent network
- 1D diverging network
- 1D small-world network
- 1D testing cell models network
- 1D testing STDP model network
- bio-realistic neural network
- large scale computation
- model replicaton (i.e., reproduce results from paper)
- parameters explorations
- basis of spiking neural network theory (https://neuronaldynamics.epfl.ch/online/index.html) or (https://neuromatch.io/academy/)
- familiarity with physical quantities related to electric circuits (e.g., voltages, conductances, currents, etc)
- basic python coding (numpy, work with dictionaries, some matplotlib tools, etc)
- import the simulator
- setup the simulator
- decide the cell types
- design the populations
- define the synapse types
- select the connection algorithm
- make the projections
- idealize the stimulus
- run the simulation
- save the results
- recover the results
- postprocessing (visualization, statistics, etc)
- close the simulations