Learning What and Where to Transfer
This addresses the challenge of applying transfer learning to real-world problems with insufficient data, particularly for researchers and practitioners dealing with heterogeneous models, though it is incremental in automating existing processes.
The paper tackles the problem of inefficient manual tuning in transfer learning between heterogeneous architectures and tasks by proposing a meta-learning approach that automatically learns what knowledge to transfer and where in the target network. The result is a method that significantly outperforms prior baselines on various datasets and network architectures.
As the application of deep learning has expanded to real-world problems with insufficient volume of training data, transfer learning recently has gained much attention as means of improving the performance in such small-data regime. However, when existing methods are applied between heterogeneous architectures and tasks, it becomes more important to manage their detailed configurations and often requires exhaustive tuning on them for the desired performance. To address the issue, we propose a novel transfer learning approach based on meta-learning that can automatically learn what knowledge to transfer from the source network to where in the target network. Given source and target networks, we propose an efficient training scheme to learn meta-networks that decide (a) which pairs of layers between the source and target networks should be matched for knowledge transfer and (b) which features and how much knowledge from each feature should be transferred. We validate our meta-transfer approach against recent transfer learning methods on various datasets and network architectures, on which our automated scheme significantly outperforms the prior baselines that find "what and where to transfer" in a hand-crafted manner.