On the Hardness of Unsupervised Domain Adaptation: Optimal Learners and Information-Theoretic Perspective

Open in new window