Understanding Self-Training for Gradual Domain Adaptation

Abstract

Machine learning systems must adapt to data distributions that evolve over time, in applications ranging from sensor networks and self-driving car perception modules to brain-machine interfaces. We consider gradual domain adaptation, where the goal is to adapt an initial classifier trained on a source domain given only unlabeled data that shifts gradually in distribution towards a target domain. We prove the first non-vacuous upper bound on the error of self-training with gradual shifts, under settings where directly adapting to the target domain can result in unbounded error. The theoretical analysis leads to algorithmic insights, highlighting that regularization and label sharpening are essential even when we have infinite data, and suggesting that self-training works particularly well for shifts with small Wasserstein-infinity distance. Leveraging the gradual shift structure leads to higher accuracies on a rotating MNIST dataset and a realistic Portraits dataset.

References in corpus (7)

• Batch Normalization: Accelerating Deep Network Training by Reducing Internal Covariate Shift
• Deep Domain Confusion: Maximizing for Domain Invariance
• FixMatch: Simplifying Semi-Supervised Learning with Consistency and Confidence
• Domain Adaptation: Learning Bounds and Algorithms
• On Learning Invariant Representation for Domain Adaptation
• Generalization error bounds in semi-supervised classification under the cluster assumption
• Understanding and Mitigating the Tradeoff Between Robustness and Accuracy

Cited by in corpus (5)

• Self-Training: A Survey
• DAugNet: Unsupervised, Multi-source, Multi-target, and Life-long Domain Adaptation for Semantic Segmentation of Satellite Images
• Adaptive Self-training for Few-shot Neural Sequence Labeling
• Self-training Avoids Using Spurious Features Under Domain Shift
• Gradual Domain Adaptation in the Wild:When Intermediate Distributions are Absent