/ReliableDosageCombi

Implementation of "Reliable Off-Policy Learning for Dosage Combinations"

Primary LanguagePythonMIT LicenseMIT

Reliable Off-Policy Learning for Dosage Combinations

Implementation of "Reliable Off-Policy Learning for Dosage Combinations".

Install requirements:

pip install -r requirements.txt

To run the experiments on the TCGA dataset:

  1. Download the TCGA dataset version from https://github.com/ioanabica/SCIGAN and store it in "dataset/tcga/".
  2. Specify the parameters and model options in "conf/config.yaml".
  3. Run the run.sh script with the following command:
bash run.sh

The results are then stored in "logs/" in an own subfolder per each setting.