Skip to content

computational efficacy of a modular spiking neural network as a function of heterogeneity of delay in the network

Notifications You must be signed in to change notification settings

moorugi98/hetero_delay_network

Repository files navigation

hetero_delay_network

Simulate a biologically plausible feed-forward modular network of Brunel networks (Brunel, 2000) using NEST-simulator. By default, the membrane potential and the spikes are saved as data in a given PATH. Measure basic properties of networks such as pairwise synchrony, revised local variation, firing rate and Fano facor. Train a ridge classifier using a state matrix of the network to test the classification ability of the network. Delay distributions can be altered depending on user's need.

  • params.py: define all the necessary parameters for the simulation.
  • main.py: script to simulate a network save data to file.
  • helpers.py: script which entails all necessary functions for data analysis.
  • measures.py: script to compute useful properties of the network. The data is saved as DataFrame.
  • train.py: train a ridge classifier using a state matrix of the network.
  • plotter.py: script to plot the result.