Federated Learning on MNIST data with attack mitigation/detection

This is a CNN model that has been trained on the MNIST data set in three levels:

• Level 1:
  • Centralized learning, no attack detection/mitigation
• Level 2:
  • Federated Learning, no attack detection/mitigation
• Level 3:
  • Federated Learning with attack detection/mitigation

How to Run

• Create a python vertual environment with conda or pip
• Check your CUDA version on your system:

nvidia-smi

• Ensure that you have pytorch install from the pytorch Official Website, copy the command according to CUDA version for example:

conda install pytorch torchvision torchaudio pytorch-cuda=12.4 -c pytorch -c nvidia

or

pip install pytorch torchvision torchaudio pytorch-cuda=12.4 -c pytorch -c nvidia

• After pytorch is installed, you will be able to run any of the models via:

py level_x.py

This will train, evaluate, and save the best model.