Skip to content

Latest commit

 

History

76 Commits

Folders and files

NameName
Last commit message
Last commit date
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

Estimating Synaptic Conductances By Learning Neuron Dynamics

Neurons are sophisticated IO machines which allow extraditing their relevant dynamics through various approaches. Since the adoption of now wide-spread machine learning methods, there has been a great interest in drwaing parallels between biological and artificial units of computation. Here we consider 4 different single neuron models (Moris-Lecar 2D, FitzHugh-Nagumo 2D, FitzHugh-Nagumo-Rinzel 3D, Wang 4D) and utilize 3 different neural architectures (Wavenet, Transformer, CNN) in portraying these models dynamics.

System Requirements

The bulk of the training and testing in this work is carried out on a NVIDIA 4070 GPU using pytorch library.

ODE dynamics of neuronal models are investigated using xpp-auto.

Neural Data and Mathematical Models

Training

Benchmarking

Citing

If you find this work useful, please consider citing it.

@article{forgione2023from,
  author={Forgione, Marco and Pura, Filippo and Piga, Dario},
  journal={IEEE Control Systems Letters}, 
  title={From System Models to Class Models:
   An In-Context Learning Paradigm}, 
  year={2023},
  volume={7},
  number={},
  pages={3513-3518},
  doi={10.1109/LCSYS.2023.3335036}
}

License

This repository is released under the MIT license. See LICENSE for additional details.

About

Project adapting various neural architectures for single neuron models and population models for neuronal coherence / voltage mapping

Topics

Resources

Stars

2 stars

Watchers

1 watching

Forks

Releases

Packages

Contributors

Languages