Predict Finger Movement

In this project we implemented different deep learning models to predict the finger movements from Electroencephalography recordings. Then, we compared them with some common baselines.

Files:

• modelWrapper.py: contains a superclass implementing the general functions adopted by all the proposed models, e.g. fit, cross-validation, score functions.
• models.py: contains the implementation of the models.
• callbacks.py: callbacks functions that can be passed to the fit() function of the models.
• test.py: trains the best model we found and shows both the train and test accuracies.
• dlc_bci.py: loads the dataset.
• helpers.py: support functions.
• Report.pdf: report which explains the problem and our approach to it.

We suggest to read the report for a detalied description.