Submodular Meta-Learning

Arman Adibi, Aryan Mokhtari, Hamed Hassani

In this paper, we introduce a discrete variant of the meta-learning framework. Meta-learning aims at exploiting prior experience and data to improve performance on future tasks. By…