/Android-Malware-Detection-Adversarial-Examples

Evaluation of ML models in Android malware classification, adversarial attacks on DNNs & defense mechanisms

Primary LanguagePython

Research on Android malware detection based on ML models & the weakness of DNNs in adversarial examples.

The experiments are carried through the Drebin dataset.

  1. preprocessing folder contains scripts related to dataset and its features.
  2. feature_based_original_dataset folder contains scripts for the experiments in whole feature space.
  3. feature_based_reduced_dataset folder contains scripts for the experiments on the reduced feature space.

Each folder contains a README file to help you through the experiments.