We provide a first-order oracle complexity lower bound for finding stationary points of min-max optimization problems where the objective function is smooth, nonconvex in the minimization variable, and strongly concave in the maximization variable. We establish a lower bound of $\Omega\left(\sqrt{\kappa}\epsilon^{-2}\right)$ for deterministic oracles, where $\epsilon$ defines the level of approximate stationarity and $\kappa$ is the condition number. Our analysis shows that the upper bound achieved in (Lin et al., 2020b) is optimal in the $\epsilon$ and $\kappa$ dependence up to logarithmic factors. For stochastic oracles, we provide a lower bound of $\Omega\left(\sqrt{\kappa}\epsilon^{-2} + \kappa^{1/3}\epsilon^{-4}\right)$. It suggests that there is a significant gap between the upper bound $\mathcal{O}(\kappa^3 \epsilon^{-4})$ in (Lin et al., 2020a) and our lower bound in the condition number dependence.
翻译:我们为寻找最小最大优化问题的固定点提供了一级或一级复杂度, 其目标功能是平滑的, 最小化变量中不是隐蔽的, 且在最大化变量中具有很强的共性。 我们为确定性( 确定性) 或甲骨蜡( 确定性) 设定了一个较低范围的 $Omega\ left (sqrt\ kappa\ eepsilon ⁇ 2 right), $\ epsilon=1/3 ⁇ epsilon ⁇ -4 ⁇ right) 。 我们的分析显示, 在( Lin等人等人, 2020b) 中实现的上限值在 $\ epsilon $ 和 $\ kapppa$ 上限值达到最佳, 直至对对对正数。 对于分析性或甲骨骼, 我们提供了较低约束 $\\ rftleft left (r) +\ kappapapaca_ 1/ {1/3 ⁇ 4 right) 和我们2020年的上限 (mathal=lusisal) asima) 3\\\\\\ lisalisa) esta) a. (\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\\