Skip to content
master
Switch branches/tags
Code

Latest commit

 

Git stats

Files

Permalink
Failed to load latest commit information.
Type
Name
Latest commit message
Commit time
 
 
 
 
 
 
 
 
 
 
 
 

CRNN-malware-detector

This is the repository for our deep learning based malware detection model. Please read the paper for any details of our research(https://doi.org/10.1016/j.ins.2020.05.026). One thing you may concern, these source codes are operatable for the fully runable 32bit PE files which are not applied any preprocessings. Also the dataset used for our experiments is consist of 1,000 malwares and benign files, respectively. However it is extremly dangerous to release fully workable malware so we uploaded only the file names or SHA1 hash values for the binary file in dataset.

Contributors

  • Seungho Jeon
  • Jongsub Moon

Dependencies

  Python3.6
  numpy-1.16.2
  tensorfllow-1.13.1
  scikit-learn-0.20.3
  capstone-4.0.1
  pefile-2019.4.18

Usages

This is the core source codes for our research. You may only concern main.py to run the our deep learning based malware detection model. Chaning some parameters written in main.py, you can train or test the our mdoel.
belows are the part of main.py

        ...
        for strategy in ['noguided', 'random', 'prob_f', 'prob_b']:
            DetectionModel.run(n_sampled_path,
                               'train',
                               '.\\indexed-paths', # data root,
                               strategy,
        ...

Put 'train' in string type for the second parameter of DetectionModel.run to train the OCAE and DRNN, 'train_lc' to train the linear classifier applying after DRNN, or 'test' to test the performance of our model.
For your convinience, we uploaded the trained model parameters so you can skip the training part :)

About

Convolutional Recurrent Neural Net based malware detection model

Resources

Releases

No releases published

Packages

No packages published

Languages