论文标题
未标记的数据帮助：minimax分析和对抗性鲁棒性
Unlabeled Data Help: Minimax Analysis and Adversarial Robustness
论文作者
论文摘要
最近提出的自我监督学习（SSL）方法成功地证明了使用其他未标记的数据补充学习算法的巨大潜力。但是，目前尚不清楚现有的SSL算法是否可以完全利用标记和未标记数据的信息。本文为基于重建的SSL算法\ Citep {Lee2020 Presspricting}提供了肯定的答案。尽管现有文献仅着重于建立收敛速率的上限，但我们提供了严格的最小值分析，并成功证明了基于重建的SSL算法在不同的数据生成模型下的速率优先性。此外，我们将基于重建的SSL纳入现有的对抗训练算法，并表明从未标记的数据中学习有助于提高鲁棒性。
The recent proposed self-supervised learning (SSL) approaches successfully demonstrate the great potential of supplementing learning algorithms with additional unlabeled data. However, it is still unclear whether the existing SSL algorithms can fully utilize the information of both labelled and unlabeled data. This paper gives an affirmative answer for the reconstruction-based SSL algorithm \citep{lee2020predicting} under several statistical models. While existing literature only focuses on establishing the upper bound of the convergence rate, we provide a rigorous minimax analysis, and successfully justify the rate-optimality of the reconstruction-based SSL algorithm under different data generation models. Furthermore, we incorporate the reconstruction-based SSL into the existing adversarial training algorithms and show that learning from unlabeled data helps improve the robustness.
