Tikhonov Regularization is Optimal Transport Robust under Martingale Constraints

Tikhonov Regularization is Optimal Transport Robust under Martingale Constraints
复制标题

Tikhonov 正则化是鞅约束下的最优传输鲁棒性

DOI:
10.48550/arxiv.2210.01413
复制
发表时间:
2022
期刊:
ArXiv
影响因子:
--
通讯作者:
Viet Anh Nguyen
Viet Anh Nguyen
中科院分区:
--
文献类型:
--
作者:
Jiajin Li;Si;J. Blanchet;Viet Anh Nguyen

文献摘要

参考文献

相似文献

分布式鲁棒优化已被证明提供了一种规范学习模型的原则性方法。在本文中,我们发现吉洪诺夫正则化在最优传输意义上具有分布鲁棒性(即,如果对手选择经验测量的合适最优传输邻域中的分布),前提是还施加了合适的鞅约束。此外,我们引入了鞅约束的放宽,这不仅为一类现有的鲁棒方法提供了统一的观点,而且还带来了新的正则化工具。为了实现这些新颖的工具,提出了易于处理的计算算法。作为副产品,本文证明的强对偶定理可以潜在地应用于其他独立感兴趣的问题。
Distributionally robust optimization has been shown to offer a principled way to regularize learning models. In this paper, we find that Tikhonov regularization is distributionally robust in an optimal transport sense (i.e., if an adversary chooses distributions in a suitable optimal transport neighborhood of the empirical measure), provided that suitable martingale constraints are also imposed. Further, we introduce a relaxation of the martingale constraints which not only provides a unified viewpoint to a class of existing robust methods but also leads to new regularization tools. To realize these novel tools, tractable computational algorithms are proposed. As a byproduct, the strong duality theorem proved in this paper can be potentially applied to other problems of independent interest.
DOI: 10.1017/jpr.2019.49
发表时间: 2019-09-01
影响因子: 1
作者:
Blanchet, Jose;Kang, Yang;Murthy, Karthyek
通讯作者: Murthy, Karthyek
DOI: --
发表时间: 2020-02
期刊: ArXiv
影响因子: --
作者:
Zac Cranko;Zhan Shi;Xinhua Zhang;R. Nock;Simon Kornblith
通讯作者: Zac Cranko;Zhan Shi;Xinhua Zhang;R. Nock;Simon Kornblith
DOI: 10.1287/moor.2021.1178
发表时间: 2018-10
期刊: Math. Oper. Res.
影响因子: --
作者:
J. Blanchet;Karthyek Murthy;Fan Zhang
通讯作者: J. Blanchet;Karthyek Murthy;Fan Zhang