Nov 11, 2026

2026-2027 Graduate Academic Calendar

MITS 6810G - Adversarial Machine Learning

This course introduces adversarial attacks and defenses against machine learning models. Covered topics include evasion attacks against learning-based schemes, causative attacks by perturbing training datasets, and an introduction to robust statistics. Additionally, the course provides an overview of attacks against learning-based schemes utilized in some of the cybersecurity applications, such as Spam Detection and Intrusion Detection. The course provides the latest overview of state-of-the-art Adverserial Machine Learning (AML) schemes, such as Generative Adversarial Networks (GAN) and Adversarial Active Learning.
Credit hours: 3
Prerequisite(s): MITS 6800G or CSCI 5770G
Experiential learning Yes