Gradient-Based Transfer Learning

Gustaf Tegnér, Alfredo Reichlin, Hang Yin, Hedvig Kjellstrom, Mårten Björkman, Danica Kragic

Keywords: meta-learning, gradient-based meta-learning, transfer learning, representation learning

TL;DR: We formulate transfer learning as a meta-learning problem and extend current gradient-based meta-learning methods to this setting.

Abstract: We formulate transfer learning as a meta-learning problem by extending upon the current meta-learning paradigm in that support and query data are drawn from different, but related distributions of tasks. Inspired by the success of Gradient-Based Meta-Learning (GBML), we propose to expand it to the transfer learning setting by constructing a general encoder-decoder architecture that learns a map between functionals of different domains. This is achieved by leveraging on the idea that the task-adapted parameters of a meta-learner can serve as an informative representation of the task itself. We demonstrate the proposed method on regression, prediction of dynamical systems and meta-imitation learning problems.

Anonymous Url: I certify that there is no URL (e.g., github page) that could be used to find authors’ identity.

No Acknowledgement Section: I certify that there is no acknowledgement section in this submission for double blind review.

Code Of Ethics: I acknowledge that I and all co-authors of this work have read and commit to adhering to the ICLR Code of Ethics

Submission Guidelines: Yes

Please Choose The Closest Area That Your Submission Falls Into: Deep Learning and representational learning