Understanding Gradual Domain Adaptation: Improved Analysis, Optimal Path and Beyond

Haoxiang Wang, Bo Li, Han Zhao

Research output: Contribution to journalConference articlepeer-review

Abstract

The vast majority of existing algorithms for unsupervised domain adaptation (UDA) focus on adapting from a labeled source domain to an unlabeled target domain directly in a one-off way. Gradual domain adaptation (GDA), on the other hand, assumes a path of (T−1) unlabeled intermediate domains bridging the source and target, and aims to provide better generalization in the target domain by leveraging the intermediate ones. Under certain assumptions, Kumar et al. (2020) proposed a simple algorithm, gradual self-training, along with a generalization bound in the order of eO(T)0+O(qlognT )) for the target domain error, where ε0 is the source domain error and n is the data size of each domain. Due to the exponential factor, this upper bound becomes vacuous when T is only moderately large. In this work, we analyze gradual self-training under more general and relaxed assumptions, and prove a significantly improved generalization bound as (Eqaution presented), where ∆ is the average distributional distance between consecutive domains. Compared with the existing bound with an exponential dependency on T as a multiplicative factor, our bound only depends on T linearly and additively. Perhaps more interestingly, our result implies the existence of an optimal choice of T that minimizes the generalization error, and it also naturally suggests an optimal way to construct the path of intermediate domains so as to minimize the accumulative path length T∆ between the source and target. To corroborate the implications of our theory, we examine gradual self-training on multiple semi-synthetic and real datasets, which confirms our findings. We believe our insights provide a path forward toward the design of future GDA algorithms.

Original languageEnglish (US)
Pages (from-to)22784-22801
Number of pages18
JournalProceedings of Machine Learning Research
Volume162
StatePublished - 2022
Event39th International Conference on Machine Learning, ICML 2022 - Baltimore, United States
Duration: Jul 17 2022Jul 23 2022

ASJC Scopus subject areas

  • Artificial Intelligence
  • Software
  • Control and Systems Engineering
  • Statistics and Probability

Fingerprint

Dive into the research topics of 'Understanding Gradual Domain Adaptation: Improved Analysis, Optimal Path and Beyond'. Together they form a unique fingerprint.

Cite this