arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2301.08015cs.GTcs.LGmath.OCstat.ML
收藏

非凸多人博弈的全局纳什均衡:理论与算法

Global Nash Equilibrium in Non-convex Multi-player Game: Theory and Algorithms

JD Explore Academy, JD.com Inc(京东集团京东探索研究院);Academy of Mathematics and Systems Science, Chinese Academy of Sciences(中国科学院数学与系统科学研究院);Research Institute for Intelligent Autonomous Systems, Tongji University(同济大学智能自主系统研究院);Systems Research Institute, Polish Academy of Sciences(波兰科学院系统研究所);AGH University of Science and Technology(AGH科技大学)

Guanpu Chen, Gehui Xu, Fengxiang He, Yiguang Hong, Leszek Rutkowski, Dacheng Tao

摘要:

广泛的机器学习任务可以被表述为非凸多人游戏,其中纳什均衡(NE)是所有玩家都可以接受的解决方案,因为没有人可以从单方面改变策略中受益。由于非凸性,获得全局NE的存在条件是具有挑战性的,更不用说设计理论上有保证的实现算法。对非凸多人对策的形式进行共轭变换,将互补问题转化为具有连续伪梯度映射的变分不等式问题.然后我们证明了整体NE的存在条件:VI问题的解满足一个对偶关系。基于此VI公式,我们设计了一个基于共轭的常微分方程(ODE)来逼近全局NE,并证明了它具有指数收敛速度。为了使动力学更容易实现,我们进一步推导出一个离散化的算法。我们将我们的算法应用到两个典型的场景:多人广义单调博弈和多人势博弈。在这两个设置中,我们证明了步长设置需要$\mathcal{O}(1/k)$和$\mathcal{O}(1/\sqrt k)$,分别产生$\mathcal{O}(1/ k)$和$\mathcal{O}(1/\sqrt k)$的收敛速度。在鲁棒神经网络训练和传感器定位中的广泛实验与我们的理论完全一致。

英文摘要:

Wide machine learning tasks can be formulated as non-convex multi-player games, where Nash equilibrium (NE) is an acceptable solution to all players, since no one can benefit from changing its strategy unilaterally. Attributed to the non-convexity, obtaining the existence condition of global NE is challenging, let alone designing theoretically guaranteed realization algorithms. This paper takes conjugate transformation to the formulation of non-convex multi-player games, and casts the complementary problem into a variational inequality (VI) problem with a continuous pseudo-gradient mapping. We then prove the existence condition of global NE: the solution to the VI problem satisfies a duality relation. Based on this VI formulation, we design a conjugate-based ordinary differential equation (ODE) to approach global NE, which is proved to have an exponential convergence rate. To make the dynamics more implementable, we further derive a discretized algorithm. We apply our algorithm to two typical scenarios: multi-player generalized monotone game and multi-player potential game. In the two settings, we prove that the step-size setting is required to be $\mathcal{O}(1/k)$ and $\mathcal{O}(1/\sqrt k)$ to yield the convergence rates of $\mathcal{O}(1/ k)$ and $\mathcal{O}(1/\sqrt k)$, respectively. Extensive experiments in robust neural network training and sensor localization are in full agreement with our theory.

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。

↑