Deep into The Domain Shift: Transfer Learning through Dependence Regularization
This work addresses domain shift issues in business and financial applications where labeling functions have different sensitivities to changes in marginals versus dependence structures, offering a more discriminative approach than existing methods.
The paper tackles the problem of domain adaptation by separately measuring differences in marginal distributions and dependence structures, rather than overall distributional discrepancies, to improve transfer learning. Experiments on three real-world datasets show notable and robust improvements over benchmark models.
Classical Domain Adaptation methods acquire transferability by regularizing the overall distributional discrepancies between features in the source domain (labeled) and features in the target domain (unlabeled). They often do not differentiate whether the domain differences come from the marginals or the dependence structures. In many business and financial applications, the labeling function usually has different sensitivities to the changes in the marginals versus changes in the dependence structures. Measuring the overall distributional differences will not be discriminative enough in acquiring transferability. Without the needed structural resolution, the learned transfer is less optimal. This paper proposes a new domain adaptation approach in which one can measure the differences in the internal dependence structure separately from those in the marginals. By optimizing the relative weights among them, the new regularization strategy greatly relaxes the rigidness of the existing approaches. It allows a learning machine to pay special attention to places where the differences matter the most. Experiments on three real-world datasets show that the improvements are quite notable and robust compared to various benchmark domain adaptation models.