This project, written by Rodrigue De Schatezen, Youssef Farag, and Philippe Solodov, represents our submission for the COVID-19 challenge of the CPSC 340 2020 Winter Term 2 final exam at UBC.
This framework can train a neural network COVID classifier, given an appropriate dataset of X-Ray images. The aim of the project was to use machine learning techniques learned throughout the course (and beyond) to design an algorithm for classifying chest X-Rays with COVID-19.
The write-up, explaining the task and our approach, can be found in project_report.pdf. The data, comprised of a train set, train labels, test set, can be found here. Also included in the data.zip is our own final_predictions.csv which are the predicted labels for the test set that achieved an F-score of 0.96551.