Primal-dual Optimization: Algorithms and Convergence

This lecture by the instructor covers primal-dual optimization algorithms for convex-concave minimax problems. It delves into the proximal point method, extra-gradient algorithm, and optimistic gradient descent ascent, discussing their convergence properties and applications. The presentation also includes comparisons of convergence rates for smooth convex-concave minimax optimization. The lecture concludes with a focus on stochastic primal-dual hybrid gradient methods for composite minimization, providing insights into their implementation and effectiveness.