You will be evaluated based on the following:
Competition test data accuracy on the Kaggle leader board.
Accuracy obtained by  code submission run on private test data by the instructor and team.