Demo for Scipy 2019
Run the task.py file first. This produces the features and saves them as pickled files
Run the train_features.py second. This gives options to run TPOT to find the optimal classifier. If not specified, the program defaults to training a RandomForest classifier
Review the argparse help strings to ensure all arguments are being passed.
You will need to download the DTI data from ppmi-info.org. If needed data can be provided upon request to reproduce results.