Distributed Variable Sample-Size Gradient-response and Best-response Schemes for Stochastic Nash Equilibrium Problems over Graphs

TL;DR

This paper develops and analyzes distributed stochastic gradient and best-response algorithms for Nash equilibrium problems over graphs, providing optimal convergence rates, complexity bounds, and communication efficiency under various assumptions.

Contribution

It introduces variable sample-size schemes for stochastic Nash games, establishing their convergence rates, complexity bounds, and extending them to distributed settings with communication efficiency.

Findings

Optimal rate statements and oracle complexity bounds derived.

Distributed protocols achieve similar complexity with reduced communication.

Numerical results support theoretical convergence and complexity claims.

Abstract

This paper considers a stochastic Nash game in which each player minimizes an expectation valued composite objective. We make the following contributions. (I) Under suitable monotonicity assumptions on the concatenated gradient map, we derive optimal rate statements and oracle complexity bounds for the proposed variable sample-size proximal stochastic gradient-response (VS-PGR) scheme when the sample-size increases at a geometric rate. If the sample-size increases at a polynomial rate of degree $v > 0$, the mean-squared errordecays at a corresponding polynomial rate while the iteration and oracle complexities to obtain an $\epsilon$-NE are $\mathcal{O}(1/\epsilon^{1/v})$ and $\mathcal{O}(1/\epsilon^{1+1/v})$, respectively. (II) We then overlay (VS-PGR) with a consensus phase with a view towards developing distributed protocols for aggregative stochastic Nash games. In the resulting scheme, when the sample-size and the consensus steps grow at a geometric and linear rate, computing an $\epsilon$-NE requires similar iteration and oracle complexities to (VS-PGR) with a communication complexity of $\mathcal{O}(\ln^2(1/\epsilon))$; (III) Under a suitable contractive property associated with the proximal best-response (BR) map, we design a variable sample-size proximal BR (VS-PBR) scheme, where each player solves a sample-average BR problem. Akin to (I), we also give the rate statements, oracle and iteration complexity bounds. (IV) Akin to (II), the distributed variant achieves similar iteration and oracle complexities to the centralized (VS-PBR) with a communication complexity of $\mathcal{O}(\ln^2(1/\epsilon))$ when the communication rounds per iteration increase at a linear rate. Finally, we present some preliminary numerics to provide empirical support for the rate and complexity statements.