arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2006.02067math.OC
收藏

随机鞍点问题的广义界

Generalization Bounds for Stochastic Saddle Point Problems

Junyu Zhang, Mingyi Hong, Mengdi Wang, Shuzhong Zhang

摘要:

研究随机鞍点(SSP)问题经验鞍点(ESP)解的广义界。对于具有Lipschitz连续和强凸-强凹目标函数的SSP,我们利用一致稳定性论证建立了一个$\mathcal{O}(1/n)$推广界.我们还提供了各种假设下的推广界,包括没有强凸性和没有有界域的情况。我们用两个例子来说明我们的结果:马尔可夫决策过程中的批量策略学习,以及随机对策的混合策略纳什均衡估计。在这些例子中,我们证明了正则化ESP解决方案具有接近最优的样本复杂度。据我们所知,这是ESP的推广理论的第一组结果。

英文摘要:

This paper studies the generalization bounds for the empirical saddle point (ESP) solution to stochastic saddle point (SSP) problems. For SSP with Lipschitz continuous and strongly convex-strongly concave objective functions, we establish an $\mathcal{O}(1/n)$ generalization bound by using a uniform stability argument. We also provide generalization bounds under a variety of assumptions, including the cases without strong convexity and without bounded domains. We illustrate our results in two examples: batch policy learning in Markov decision process, and mixed strategy Nash equilibrium estimation for stochastic games. In each of these examples, we show that a regularized ESP solution enjoys a near-optimal sample complexity. To the best of our knowledge, this is the first set of results on the generalization theory of ESP.

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。

↑