This repository contains our work on the first project of the EPFL Deep Learning course. The goal was to perform a binary classification on EEG data to predict the laterality of finger movements.
A Neural Network was designed from scratch to perform this task. The used framework was PyTorch. Moreover the main code implements different baselines using the scikit-learn library.
The main executable script is the test.py file. By running it, the dataset will be loaded if already installed. If not, it will be downloaded using the dlc_bci.py script. The file contains different functions to compute a baseline for the task using an LDA and SVM classifier. All different informations on the implementation and development of the network is explained in the report.