This repo contains all a framework for training a classifier to classify ddos-attacks on the marist ddos attack dataset. Also contained in this work is ability to train a classifer that is robust to common gradient based (Gradient Sign Attack and Projected Gradient Decent Attack) and gradient free(Single Pixel Attack and Jacobian Saliency Map Attack) adversarial attacks by using Robust Adversarial Training as outlined in TOWARDS DEEP LEARNING MODELS RESISTANT TO ADVERSARIAL ATTACKS.

To run all experiments (5 replicas per experiment) please find exp_runner.sh.