Anomaly Detection of Unknown Malware Families

A model for anomaly detection of unknown malware families using features extracted from a deep learning model trained on known malware family samples.

Structure

  • main.py : Core of the Program
  • embedding_train.py : Training the embedding model
  • anomaly_train.py : Training the anomaly model
  • test.py : Test processing for multi-class classification tasks and anomaly detection tasks
  • visualization.py : Visualization process by GUI
  • preprocessing.py : Preprocessing of input data
  • config.py : Manage global variables used throughout the program
  • parse_args.py : Process to receive arguments by command line
  • models
    • alex_net.py
    • efficient_net.py
    • metrics.py