This repository contains the C/C++ source codes, PyNN and PyNEST scripts developed as part of the following studies:
Trensch, G., Gutzen, R., Blundell, I., Denker, M., and Morrison, A. (2018). Rigorous neural network simulations: A model substantiation methodology for increasing the correctness of simulation results in the absence of experimental validation data. Frontiers in Neuroinformatics 12, 81. doi:10.3389/fninf. 2018.00081
Gutzen, R., von Papen, M., Trensch, G., Quaglio, P., Grün, S., and Denker, M. (2018). Reproducible neural network simulations: Statistical methods for model validation on the level of network activity data. Frontiers in Neuroinformatics 12, 90. doi:10.3389/fninf.2018.00090
Trensch, G., and Morrison, A. (2022). A System-on-Chip Based Hybrid Neuromorphic Compute Node Architecture for Reproducible Hyper-Real-Time Simulations of Spiking Neural Networks. Frontiers in Neuroinformatics 16:884033.doi: 10.3389/fninf.2022.884033
- Console application to evaluate different ODE solver strategies for solving the Izhikevich neuron model dynamics. (source)
- Implementations of the two-population Izhikevich network model described below:
The codes were developed to explore different ODE solver strategies and to determine the required numerical precision needed to capture the dynamics of the Izhikevich neuron model [2] with sufficient accuracy. The two-population Izhikevich network model described below and originally published in [3], was used for a quantitative assessment of different implementations, namely: a reference implementation in the C language, a SpiNNaker PyNN implementation, an implementation using the NEST simulation tool, and an implementation on a novel neuromorphic compute node architecture. The model was also used to perform benchmarking tasks.
The Izhikevich neuron model was originally published in [2] and the two-population network model in [3]. The tables (Tab. 1, 2) below summarize the properties and parameters of the network, following the proposed methods described in [4] and [6]. In order to avoid the occurrence of simulation artifacts, the temporal resolution of the simulation is set to 0.1 ms. This is a 10 times smaller value than used by the original implementation [3]. See also [5], in which the reproducibility of the two-population Izhikevich network model was evaluated using the NEST simulator.
[1] Gutzen, R., von Papen, M., Trensch, G., Quaglio, P., Grün, S., and Denker, M. (2018). Reproducible neural network simulations: Statistical methods for model validation on the level of network activity data. Frontiers in Neuroinformatics 12, 90. doi:10.3389/fninf.2018.00090
[2] Izhikevich, E. M. (2003). Simple model of spiking neurons. Trans. Neur. Netw., 14(6):1569–1572.
[3] Izhikevich, E. M. (2006). Polychronization: Computation with spikes. Neural Computation, 18:245–282.
[4] Nordlie, E., Gewaltig, M.-O., and Plesser, H. E. (2009). Towards Reproducible Descriptions of Neuronal Network Models. PLoS Computational Biology, 5(8):e1000456.
[5] Pauli, R., Weidel, P., Kunkel, S., and Morrison, A. (2018). Reproducing polychronization: A guide to maximizing the reproducibility of spiking network models. Frontiers in Neuroinformatics 12. doi:10.3389/fninf.2018.00046
[6] Senk, J., Kriener, B., Djurfeldt, M., Voges, N., Jiang, H.-J., Schüttler, L., Gramelsberger, G., Diesmann, M., Plesser, H. E., and van Albada, S. J. (2021). Connectivity concepts in neuronal network modeling.
[7] Trensch, G., Gutzen, R., Blundell, I., Denker, M., and Morrison, A. (2018). Rigorous neural network simulations: A model substantiation methodology for increasing the correctness of simulation results in the absence of experimental validation data. Frontiers in Neuroinformatics 12, 81. doi:10.3389/fninf. 2018.00081