Towards Backwards-Compatible Data with Confounded Domain Adaptation
This work addresses a specific bottleneck in domain adaptation for researchers and practitioners dealing with confounded shifts, offering a method to make adapted data broadly usable, though it appears incremental as it builds on generalized label shift concepts.
The paper tackles the problem of domain adaptation when covariate and label shifts occur simultaneously and are confounded, aiming to achieve general-purpose backwards compatibility for adapted data to be used in various downstream tasks. The result is a novel framework that minimizes divergence between source and target distributions, demonstrated on synthetic and real datasets.
Most current domain adaptation methods address either covariate shift or label shift, but are not applicable where they occur simultaneously and are confounded with each other. Domain adaptation approaches which do account for such confounding are designed to adapt covariates to optimally predict a particular label whose shift is confounded with covariate shift. In this paper, we instead seek to achieve general-purpose data backwards compatibility. This would allow the adapted covariates to be used for a variety of downstream problems, including on pre-existing prediction models and on data analytics tasks. To do this we consider a modification of generalized label shift (GLS), which we call confounded shift. We present a novel framework for this problem, based on minimizing the expected divergence between the source and target conditional distributions, conditioning on possible confounders. Within this framework, we provide concrete implementations using the Gaussian reverse Kullback-Leibler divergence and the maximum mean discrepancy. Finally, we demonstrate our approach on synthetic and real datasets.