f -Domain-Adversarial Learning: Theory and Algorithms

             David Acuna 1 2 3 Guojun Zhang 4 3 Marc T. Law 1 Sanja Fidler 1 2 3

             Abstract                 labeled and unlabeled datasets, and it is expected to gen-
                                 eralize well to the target dataset if the gap between both
   Unsupervised domain adaptation is used in many        domains is not very significant.
   machine learning applications where, during train-
   ing, a model has access to  ...