Аннотация:
Extragradient method (EG) is one of the most popular methods for solving saddle point and variational inequalities problems (VIP). Despite its long history and significant attention in the optimization community, there remain important open questions about convergence of EG. In this paper, we resolve one of such questions and derive the first last-iterate O(1/K) convergence rate for EG for monotone and Lipschitz VIP without any additional assumptions on the operator. The rate is given in terms of reducing the squared norm of the operator. Moreover, we establish several results on the (non-)cocoercivity of the update operators of EG, Optimistic Gradient Method, and Hamiltonian Gradient Method, when the original operator is monotone and Lipschitz.
The central part of our analysis is based on Performance Estimation Problems and computer-assistant proofs. In the talk, I will pay special attention to this approach and explain the main non-trivial issues we faced on the way of getting the final proofs via numerical computations.