Logarithmic Query Complexity for Approximate Nash Computation in Large Games

@article{Goldberg2018LogarithmicQC,
  title={Logarithmic Query Complexity for Approximate Nash Computation in Large Games},
  author={Paul W. Goldberg and Francisco Javier Marmolejo-Coss{\'i}o and Zhiwei Steven Wu},
  journal={Theory of Computing Systems},
  year={2018},
  volume={63},
  pages={26-53}
}
We investigate the problem of equilibrium computation for “large” n-player games. Large games have a Lipschitz-type property that no single player’s utility is greatly affected by any other individual player’s actions. In this paper, we mostly focus on the case where any change of strategy by a player causes other players’ payoffs to change by at most 1n$\frac {1}{n}$. We study algorithms having query access to the game’s payoff function, aiming to find ε-Nash equilibria. We seek algorithms… 
Lower Bounds for the Query Complexity of Equilibria in Lipschitz Games
TLDR
This work develops a query-efficient reduction from more general games to Lipschitz games, and provides an exponential lower bound on the deterministic query complexity of finding -approximate correlated equilibria of n-player, m-action, λ-Lipschitzer games for strong values of , motivating the consideration of explicitly randomized algorithms in the above results.
Query complexity of approximate equilibria in anonymous games
Optimally Deceiving a Learning Leader in Stackelberg Games
TLDR
This paper shows that it is always possible for the follower to efficiently compute (near-)optimal fake payoffs, for various scenarios of learning interaction between the leader and the follower and establishes an interesting connection between the follower's deception and the leader’s maximin utility.

References

SHOWING 1-10 OF 19 REFERENCES
Query Complexity of Approximate Equilibria in Anonymous Games
TLDR
It is proved that $$\varOmega n \log {n}$$ payoffs must be queried in order to find any $$epsilon $$-well-supported Nash equilibrium, even by randomized algorithms, which is the first one to obtain an inverse polynomial approximation in poly-time.
Learning equilibria of games via payoff queries
TLDR
This work studies a corresponding computational learning model, and the query complexity of learning equilibria for various classes of games, and has the stronger result that an equilibrium can be identified while only learning a small fraction of the cost values.
Finding Approximate Nash Equilibria of Bimatrix Games via Payoff Queries
TLDR
It is shown that randomized algorithms require Ω(k2) payoff queries in order to find an ϵ-Nash equilibrium with ϵ < 1/4k, even in zero-one constant-sum games, which rules out query-efficient randomized algorithms for finding exact Nash equilibria.
Well-Supported vs. Approximate Nash Equilibria: Query Complexity of Large Games
TLDR
It is proved that, for some constant $\epsilon>0$, any randomized oracle algorithm that computes an $\epSilon$-ANE in a binary-action, $n$-player game must make $2^{\Omega(n/\log n)}$ payoff queries.
Bounds for the Query Complexity of Approximate Equilibria
TLDR
The number of payoff queries needed to compute approximate equilibria of multi-player games is analyzed, and it is found that query complexity is an effective tool for distinguishing the computational difficulty of alternative solution concepts, and new techniques for upper- and lower bounding the query complexity are developed.
The Query Complexity of Correlated Equilibria
Mechanism design in large games: incentives and privacy
TLDR
The main technical result is an algorithm for computing a correlated equilibrium of a large game while satisfying joint differential privacy, which ends up satisfying a strong privacy property as well.
Best-reply dynamics in large binary-choice anonymous games
Global Nash convergence of Foster and Young's regret testing
Query Complexity of Correlated Equilibrium
TLDR
A query model in which an n-player game is specified via a black box that returns players' utilities at pure action profiles is considered, establishing that in order to compute a correlated equilibrium, any deterministic algorithm must query the black box an exponential number of times.
...
1
2
...