Please read the attached PDF for more information. I did not get relevant results and stopped working on it, and PDF is just a draft, but it can give you an idea of what i was building.
A series of experiments were done with the objective to perform binary classification of EEG data related to physical reaction tasks using a convolutional network model. Datasets used for classification correspond to a hand-squeeze experiment and a simulated driving experiment. They're both available online.Hyperparameters such as ConvNet’s number, Kernel Size,convolutional layer depth, batch size were varied. Pre-processing techniques of data to complement classification performance was included. Experiments show that this type of architecture combina- tions did not perform well, obtaining accuracy rates that did not go past .70