Skip to content

Latest commit

 

History

History
29 lines (21 loc) · 1006 Bytes

README.md

File metadata and controls

29 lines (21 loc) · 1006 Bytes

deep_retina_pytorch

Reimplementation of deep_retina in Pytorch

Paper and original code in Keras

Requirements

  • Pytorch
  • tensorboardx
  • numpy
  • matplotlib and seaborn

Results

For this results, I trained the model on the data from paper "Analysis of Neuronal Spike Trains, Deconstructed, by Aljadeff et. al. 2016" with some arbitrary hyperparameters. For better results, more hyperparameter tuning is needed.

Spike count prediction sample

Spike count prediction sample

Filters sample

Filters sample

TO DO

  • CNN implementation
  • Parametric SoftPlus
  • L1 and L2 regularization
  • Visualization
  • Recurrent NN
  • Hyperparameter tuning