Tarzanagh Davoud Ataee, Nazari Parvin, Hou Bojian, Shen Li, Balzano Laura
University of Pennsylvania.
Amirkabir University of Technology.
Proc Mach Learn Res. 2024 May;238:2854-2862.
This paper introduces an setting in which a sequence of time-varying bilevel problems is revealed one after the other. We extend the known regret bounds for single-level online algorithms to the bilevel setting. Specifically, we provide new notions of , develop an online alternating time-averaged gradient method that is capable of leveraging smoothness, and give regret bounds in terms of the path-length of the inner and outer minimizer sequences.
本文介绍了一种场景，其中一系列时变双层问题相继出现。我们将单级在线算法的已知遗憾界扩展到双层场景。具体来说，我们提出了新的概念，开发了一种能够利用平滑性的在线交替时间平均梯度方法，并根据内层和外层最小化器序列的路径长度给出遗憾界。
