A Fine-Grained Analysis on Distribution Shift
arXiv:2110.11328
Abstract
Robustness to distribution shifts is critical for deploying machine learning models in the real world. Despite this necessity, there has been little work in defining the underlying mechanisms that cause these shifts and evaluating the robustness of algorithms across multiple, different distribution shifts. To this end, we introduce a framework that enables fine-grained analysis of various distribution shifts. We provide a holistic analysis of current state-of-the-art methods by evaluating 19 distinct methods grouped into five categories across both synthetic and real-world datasets. Overall, we train more than 85K models. Our experimental framework can be easily extended to include new methods, shifts, and datasets. We find, unlike previous work~\citep{Gulrajani20}, that progress has been made over a standard ERM baseline; in particular, pretraining and augmentations (learned or heuristic) offer large gains in many cases. However, the best methods are not consistent over different datasets and shifts.
References in corpus (12)
- Learning Transferable Visual Models From Natural Language Supervision
- Learning Transferable Features with Deep Adaptation Networks
- Do ImageNet Classifiers Generalize to ImageNet?
- WILDS: A Benchmark of in-the-Wild Distribution Shifts
- Improve Unsupervised Domain Adaptation with Mixup Training
- Heterogeneous Domain Generalization via Domain Mixup
- Just Train Twice: Improving Group Robustness without Training Group Information
- Noise or Signal: The Role of Image Backgrounds in Object Recognition
- Model Patching: Closing the Subgroup Performance Gap with Data Augmentation
- Support and Invertibility in Domain-Invariant Representations
- When Unseen Domain Generalization is Unnecessary? Rethinking Data Augmentation
- Visual Representation Learning Does Not Generalize Strongly Within the Same Domain