Timezone: »

Oral
Gradient-Based Meta-Learning with Learned Layerwise Metric and Subspace
Yoonho Lee · Seungjin Choi

Fri Jul 13 07:50 AM -- 08:00 AM (PDT) @ Victoria

Gradient-based meta-learning methods leverage gradient descent to learn the commonalities among various tasks.While previous such methods have been successful in meta-learning tasks, they resort to simple gradient descent during meta-testing.Our primary contribution is the {\em MT-net}, which enables the meta-learner to learn on each layer's activation space a subspace that the task-specific learner performs gradient descent on.Additionally, a task-specific learner of an {\em MT-net} performs gradient descent with respect to a meta-learned distance metric,which warps the activation space to be more sensitive to task identity.We demonstrate that the dimension of this learned subspace reflects the complexity of the task-specific learner's adaptation task, and also that our model is less sensitive to the choice of initial learning rates than previous gradient-based meta-learning methods.Our method achieves state-of-the-art or comparable performance on few-shot classification and regression tasks.