Lecture
This lecture covers the fundamentals of adversarial machine learning, focusing on the generation of adversarial examples, robustness challenges, and techniques such as Fast Gradient Sign Method and Proximal Gradient Descent. It explores the impact of adversarial attacks on linear models and neural networks, with a connection to majorization-minimization. The lecture also delves into adversarial training, optimization problems, and the application of Danskin's theorem in the context of machine learning.