Introduction

Implementation of the experiments as submitted to the Information Sciences Journal.

Job Configurations

The job configurations for running different mutual information estimation techniques on synthetic datasets, used by cluster_script.py, are uploaded as:

1. database\mutual_information.avail_jobs.csv
2. database\mutual_information.running_jobs.csv
3. database\automl.avail_jobs.csv
4. database\automl.running_jobs.csv

For running the experiments to detect information leakage on real datasets generated using OpenSSL TLS server, used by cluster_script_ild.py, the configurations are uploaded as:

1. database\leakage_detection_padding.avail_jobs.csv
2. database\leakage_detection_padding.running_jobs.csv

These files contain all the necessary configurations to execute the respective tasks within your experiments. Ensure to update and review these files as needed for your simulations.

You can find the folder containing these files at the following link: database folder.

Installation

The package autoqild package used for running the experiments can be installed using the instructions below:

The latest release version of AutoMLQuantILDetect can be installed from GitHub as follows::

pip install git+https://github.com/LeakDetectAI/AutoMLQuantILDetect.git

Another option is to clone the repository and install AutoMLQuantILDetect using::

python setup.py install

Dependencies

AutoMLQuantILDetect depends on the following libraries:

• AutoGLuon
• TabPFN
• Pytorch
• Tensorflow
• NumPy
• SciPy
• matplotlib
• Scikit-learn
• tqdm
• pandas (required for data processing and generation)

Citing autoqild

If you use this toolkit in your research, please cite our paper available on arXiv:

@article{gupta2024information,
  title={Information Leakage Detection through Approximate Bayes-optimal Prediction},
  author={Pritha Gupta, Marcel Wever, and Eyke Hüllermeier},
  year={2024},
  eprint={2401.14283},
  archivePrefix={arXiv},
  primaryClass={stat.ML}
}