Lecture
This lecture covers the concept of projected gradient descent with a focus on quadratic penalty. It explains the assumptions required for the method, such as non-empty closed sets and Lipschitz continuity. The lecture also discusses the convergence properties of the algorithm and its optimality conditions.