论文标题
光谱风险度量的随机优化
Stochastic Optimization for Spectral Risk Measures
论文作者
论文摘要
光谱风险目标 - 也称为$ L $ - 风险 - 允许学习系统在优化平均案例绩效（如经验风险最小化）和任务上最坏情况下的性能之间进行插值。我们开发了随机算法来通过表征其亚差异并应对诸如亚级别估计的偏见和目标的不平滑度来优化这些数量。我们在理论上和实验上表明，诸如随机亚级别和双重平均诸如偏见之类的盒子外方法受到偏见的阻碍，并且我们的方法表现优于它们。
Spectral risk objectives - also called $L$-risks - allow for learning systems to interpolate between optimizing average-case performance (as in empirical risk minimization) and worst-case performance on a task. We develop stochastic algorithms to optimize these quantities by characterizing their subdifferential and addressing challenges such as biasedness of subgradient estimates and non-smoothness of the objective. We show theoretically and experimentally that out-of-the-box approaches such as stochastic subgradient and dual averaging are hindered by bias and that our approach outperforms them.
