Efficiency through Diversity in Ensemble Models applied to Side-Channel Attacks -- A Case Study on Public-Key Algorithms --
The current repository is associated with the article "Efficiency through Diversity in Ensemble Models applied to Side-Channel Attacks" available on IACR Transactions on Cryptographic Hardware and Embedded Systems (TCHES) and the eprints
Each dataset is composed of the following scripts and repositories:
- ensemble_model.py: provides the script to use the ensembling loss introduced in the article,
- "dataset": contains the targeted dataset (should be provided by the users),
- "trained_models": contains the targeted dataset (should be provided by the users),
- "trained_models": contains the model trained by the users.
As, the considered dataset is not publicly available, the users have to perform the ensembling loss on their own datasets.
Requirements:
- Python 3.6.9
- Tensorflow 1.8.0
- Keras 2.2.4
- Scikit-learn
- Numpy