Recently, learning a model that generalizes well on out-of-distribution (OOD) data has attracted great attention in the machine learning community. In this paper, after defining OOD generalization via Wasserstein distance, we theoretically show that a model robust to input perturbation generalizes well on OOD data. Inspired by previous findings that adversarial training helps improve input-robustness, we theoretically show that adversarially trained models have converged excess risk on OOD data, and empirically verify it on both image classification and natural language understanding tasks. Besides, in the paradigm of first pre-training and then fine-tuning, we theoretically show that a pre-trained model that is more robust to input perturbation provides a better initialization for generalization on downstream OOD data. Empirically, after fine-tuning, this better-initialized model from adversarial pre-training also has better OOD generalization.
翻译:最近,在机器学习界中,学习了一种在分配外数据上十分概括的模型,引起了机器学习界的极大关注。 在本文中,在通过瓦瑟斯坦距离界定OOD一般化之后,我们理论上表明,一种对输入扰动的强型模型能够很好地概括OOD数据。 以往的研究结果认为,对抗性培训有助于改善输入-有机燃烧,我们理论上表明,经过对抗性培训的模型在OOOD数据上已经汇集了过多的风险,并在图像分类和自然语言理解任务上都进行了经验性核查。 此外,在第一次培训前和随后的微调的范例中,我们理论上表明,一种对输入扰动性更强的预培训型模型为下游OOOD数据的一般化提供了更好的初始化。 在微调后,这种经过更好的初始化的在对抗性培训前的模型也得到了更好的OOD一般化。