This is the code used for our paper: Unsupervised Deep Learning for an Image Based Network Intrusion Detection System.
Our implementations for the Wasserstein Autoencoder, Wasserstein BiGAN, and creating images can be found in this repository.
This code is intended to run in an anaconda virtual environment. To install all required packages, run the following command:
conda env create -f environment.yml
This file contains our implementation of converting feature vectors from the Bot-IoT dataset into images. A directory of images can be created by running this command:
python create_bot.py
This file will train a Wasserstein Autoencoder model and save checkpoints to the Models directory by running this command:
python wae.py
This file will train a Wasserstein BiGAN model and save checkpoints to the Models directory by running this command:
python bigan.py
This directory contains the channel vectors that are used for converting feature vectors into images.