arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2310.19035cs.LGstat.ML
收藏

通过环境增强的图不变学习是否真正学到了不变性?

Does Invariant Graph Learning via Environment Augmentation Learn Invariance?

The Chinese University of Hong Kong(香港中文大学);Tencent AI Lab(腾讯AI实验室);Hong Kong Baptist University(香港浸会大学)

Yongqiang Chen, Yatao Bian, Kaiwen Zhou, Binghui Xie, Bo Han, James Cheng

AI总结:

本文指出仅靠环境增强无法学习图不变表示,提出最小假设条件并设计GALA框架,通过助手模型识别不变子图,在DrugOOD等数据集上验证了其分布外泛化效果。

摘要:

不变图表示学习旨在从不同环境中学习数据间的不变性,以实现图上的分布外泛化。由于图环境划分通常获取代价高昂,增强环境信息已成为事实上的常用方法。然而,增强后的环境信息是否有用从未得到验证。本文发现,在没有额外假设的情况下,通过环境增强来学习不变图表示从根本上是不可能的。因此,我们提出了一组最小假设,包括变异充分性和变异一致性,以实现可行的不变图学习。随后,我们提出了一个新的框架——图不变学习助手(GALA)。GALA引入了一个需要对图环境变化或分布偏移敏感的助手模型。助手模型代理预测的正确性因此可以区分虚假子图中的变异。我们证明,在已建立的最小假设下,提取对代理预测具有最大不变性的子图,可以可证明地识别出潜在的不变子图,从而实现成功的分布外泛化。在包括DrugOOD在内的多个具有不同图分布偏移的数据集上进行的大量实验证实了GALA的有效性。

英文摘要:

Invariant graph representation learning aims to learn the invariance among data from different environments for out-of-distribution generalization on graphs. As the graph environment partitions are usually expensive to obtain, augmenting the environment information has become the de facto approach. However, the usefulness of the augmented environment information has never been verified. In this work, we find that it is fundamentally impossible to learn invariant graph representations via environment augmentation without additional assumptions. Therefore, we develop a set of minimal assumptions, including variation sufficiency and variation consistency, for feasible invariant graph learning. We then propose a new framework Graph invAriant Learning Assistant (GALA). GALA incorporates an assistant model that needs to be sensitive to graph environment changes or distribution shifts. The correctness of the proxy predictions by the assistant model hence can differentiate the variations in spurious subgraphs. We show that extracting the maximally invariant subgraph to the proxy predictions provably identifies the underlying invariant subgraph for successful OOD generalization under the established minimal assumptions. Extensive experiments on datasets including DrugOOD with various graph distribution shifts confirm the effectiveness of GALA.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑