Evaluation of uncertainty quantification to reduce False Positives and the relationship between Bayesian approach to deep learning models and their capacity

Files developed to the work here with (preprint) in here

Instructions to reproduce the results

• Download this repo in order to have the same file structure.

• Run the model

  • In order to run one experiment, run the following command:

    bash run_automatic_testCNN.sh

• Run the evalution under the defined metrics, run_metrics.py

• Run compare_methods to generate the figures in the paper