bachnguyente / echoes-mod

Machine Learning with Echo State Networks, a scikit-learn compatible package. Copied from Fabrizio Damicelli, aim to modify to accommodate other connectomes and further testing.

echoes

• Machine learning with Echo State Networks(ESN)
• High level API
• Follows scikit-learn design, eg fit-predict interface
• scikit-learn compatible, eg tools like GridSearchCV work out of the box

Quick start: See Documentation and examples.

Installation

pip install echoes

Citing

If you find echoes useful for a publication, then please use the following BibTeX to cite it:

@misc{echoes,
  author = {Damicelli, Fabrizio},
  title = {echoes: Echo State Networks with Python},
  year = {2019},
  publisher = {GitHub},
  journal = {GitHub repository},
  howpublished = {\url{https://github.com/fabridamicelli/echoes}},
}

Dependencies

• numpy
• sklearn
• pandas
• matplotlib
• seaborn
• numba
• tests:
  • mypy
  • pytest

The code has been tested with Python 3.8.12 and 3.9.7 on Ubuntu 20.04.3 LTS.

Tests

Run tests with

make test

Contributing

Any feedback on bugs, suggestions, problems? Just open an issue, I'll get back to you very soon.

References

• Reservoir computing approaches to recurrent neural network training, Mantas & Jaeger, 2009
• A Practical Guide to Applying Echo State Networks, Mantas, 2012
• Echo state network (scholarpedia), Jaeger
• Short Term Memory in Echo State Networks, Jaeger, 2001