Extragradient SVRG for Variational Inequalities: Error Bounds and Increasing Iterate Averaging
share
We study variance reduction methods for extragradient (EG) algorithms for a class of variational inequalities satisfying a classical error-bound condition. Previously, linear convergence was only known to hold under strong monotonicity. The error-bound condition is much weaker than strong monotonicity and captures a larger class of problems, including bilinear saddle-point problems such as those arising from two-player zero-sum Nash equilibrium computation. We show that EG algorithms with SVRG-style variance reduction (SVRG-EG) achieve linear convergence under the error-bound condition. In addition, motivated by the empirical success of increasing iterate averaging techniques in solving saddle-point problems, we also establish new convergence results for variance-reduced EG with increasing iterate averaging. Finally, we conduct numerical experiments to demonstrate the advantage of SVRG-EG, with and without increasing iterate averaging, over deterministic EG.
READ FULL TEXT