arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2006.07500cs.LGcs.AIstat.ML
收藏

基于因果匹配的领域泛化

Domain Generalization using Causal Matching

Microsoft Research(微软研究院)

Divyat Mahajan, Shruti Tople, Amit Sharma

摘要:

在领域泛化文献中,一个共同的目标是在类标签的条件下学习独立于领域的表示。我们表明,这个目标是不够的:存在反例,即使在满足类条件域不变性后,模型也无法推广到看不见的域。我们通过一个结构因果模型来正式说明这一观察结果,并展示了建模类内变化对泛化的重要性。具体来说,类包含表征特定因果特征的对象,域可以被解释为对这些对象的干预,改变非因果特征。我们强调一个替代条件:如果域之间的输入来自同一对象,则它们应该具有相同的表示。基于这一目标,我们提出了基于匹配的算法时,观察到的基础对象(例如,通过数据增强),并在未观察到对象时近似目标(MatchDG)。我们简单的基于匹配的算法是有竞争力的旋转MNIST,时尚MNIST,PACS和胸部X射线数据集的域外精度的先前工作。我们的方法MatchDG还恢复了地面实况对象匹配:在MNIST和Fashion-MNIST上,来自MatchDG的前10个匹配与地面实况匹配有超过50%的重叠。

英文摘要:

In the domain generalization literature, a common objective is to learn representations independent of the domain after conditioning on the class label. We show that this objective is not sufficient: there exist counter-examples where a model fails to generalize to unseen domains even after satisfying class-conditional domain invariance. We formalize this observation through a structural causal model and show the importance of modeling within-class variations for generalization. Specifically, classes contain objects that characterize specific causal features, and domains can be interpreted as interventions on these objects that change non-causal features. We highlight an alternative condition: inputs across domains should have the same representation if they are derived from the same object. Based on this objective, we propose matching-based algorithms when base objects are observed (e.g., through data augmentation) and approximate the objective when objects are not observed (MatchDG). Our simple matching-based algorithms are competitive to prior work on out-of-domain accuracy for rotated MNIST, Fashion-MNIST, PACS, and Chest-Xray datasets. Our method MatchDG also recovers ground-truth object matches: on MNIST and Fashion-MNIST, top-10 matches from MatchDG have over 50% overlap with ground-truth matches.

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。

↑