RIMS was a final project for Dr. Brian Litt's class on Brain Computer Interfaces at the University of Pennsylvania. Provided with ECoG data for 3 patients we were tasked with training models to predict finger flexion angles based on new ECoG data for each of these same patients. This repository contains all of the code used to train and make predictions with the models as well as a final report documenting how the algorithm was developed and how it performed. The training data, the models themselves, and the predictions made with them are not included here but can be provided on request.