arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 0907.1916cs.GTcs.LG
收藏

博弈中学习均衡的随机分布算法

Learning Equilibria in Games by Stochastic Distributed Algorithms

Ecole Polytechnique(巴黎综合理工学院);Centre National de la Recherche Scientifique(法国国家科学研究中心);Université de Versailles(凡尔赛大学)

Olivier Bournez, Johanne Cohen

摘要:

我们考虑一类完全随机和完全分布式的算法,我们证明学习游戏中的均衡。事实上,我们考虑一个家庭的随机分布动力学,我们证明弱收敛(在概率过程的弱收敛的意义上)对他们的平均场极限,即在一般情况下的常微分方程(ODE)。然后,我们专注于一类随机动力学,这ODE原来是与多种群复制动力学。使用已知的事实收敛的常微分方程,我们讨论的初始随机动力学的收敛性:对于一般的游戏,可能会有不收敛,但当收敛的常微分方程举行,认为随机算法收敛到纳什均衡。对于允许李雅普诺夫函数的游戏,我们称之为李雅普诺夫游戏,随机动力学收敛。我们证明了任何序数潜在的游戏,因此任何潜在的游戏是一个李雅普诺夫游戏,多仿射李雅普诺夫函数。对于具有多仿射李雅普诺夫函数的李雅普诺夫对策,证明了该李雅普诺夫函数是随机动力系统上的上鞅。这导致了一种方法来提供他们的收敛时间的界限的鞅参数。这特别适用于许多类的游戏,已被认为是在文学,包括几个负载均衡游戏场景和拥塞游戏。

英文摘要:

We consider a class of fully stochastic and fully distributed algorithms, that we prove to learn equilibria in games. Indeed, we consider a family of stochastic distributed dynamics that we prove to converge weakly (in the sense of weak convergence for probabilistic processes) towards their mean-field limit, i.e an ordinary differential equation (ODE) in the general case. We focus then on a class of stochastic dynamics where this ODE turns out to be related to multipopulation replicator dynamics. Using facts known about convergence of this ODE, we discuss the convergence of the initial stochastic dynamics: For general games, there might be non-convergence, but when convergence of the ODE holds, considered stochastic algorithms converge towards Nash equilibria. For games admitting Lyapunov functions, that we call Lyapunov games, the stochastic dynamics converge. We prove that any ordinal potential game, and hence any potential game is a Lyapunov game, with a multiaffine Lyapunov function. For Lyapunov games with a multiaffine Lyapunov function, we prove that this Lyapunov function is a super-martingale over the stochastic dynamics. This leads a way to provide bounds on their time of convergence by martingale arguments. This applies in particular for many classes of games that have been considered in literature, including several load balancing game scenarios and congestion games.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑