Tackling unsupervised multi-source domain adaptation with optimism and consistency
This work addresses a key bottleneck in domain adaptation for machine learning applications, though it appears incremental as it builds on existing single-source adaptation approaches.
The paper tackles the problem of adjusting mixture distribution weights and ensuring low error on the target domain in unsupervised multi-source domain adaptation, achieving state-of-the-art results with a novel framework using optimistic objectives and consistency regularization.
It has been known for a while that the problem of multi-source domain adaptation can be regarded as a single source domain adaptation task where the source domain corresponds to a mixture of the original source domains. Nonetheless, how to adjust the mixture distribution weights remains an open question. Moreover, most existing work on this topic focuses only on minimizing the error on the source domains and achieving domain-invariant representations, which is insufficient to ensure low error on the target domain. In this work, we present a novel framework that addresses both problems and beats the current state of the art by using a mildly optimistic objective function and consistency regularization on the target samples.