This package contains the following:
- prediction: Code and data to train integrated machine learning models to predict response to NACT.
- trained_models: Pickle files with trained models. Some files may need to be regenerated locally due to GitHub's file size constraints.
- test_results: Results obtained by applying the trained models on the holdout and external validation datasets.
- figures: Python notebooks used to generate results in the paper.