Supervised learning in spiking neural networks for precise temporal encoding
Code written in pyNN for training a single-layer, feed-forward spiking network with all-to-all connectivity to form associations between arbitrary input and target output spike patterns. Alternatively, input patterns may be associated with target output spike patterns provided by other (tutor) neurons. More details in: Gardner, B. & Grüning, A. (2016). Supervised learning in spiking neural networks for precise temporal encoding. PLoS ONE 11(8): e0161335. doi:10.1371/journal.pone.0161335.
- Python 2.7
- pyNN 0.8.1
- Numpy
- Matplotlib
- nest 2.10.0 (backend simulator used in this case)
In main_pattern_association.py: Param(1, 1, 4, 200, 1, 200) initialises a network with the following parameters:
- 1 input class
- 1 input pattern assigned to the class
- 4 target spikes assigned to each output neuron
- 200 input spike trains
- 1 output neuron
- 200 learning trials
Simulation is run with random initial network input & target spike times and weight values.
Code released under the GNU General Public License v3.