Learning-to-Learn

Learning-to-learn, also known as meta-learning, is an approach in which a system learns how to improve its own learning process across tasks. It aims to enable faster adaptation, better generalization, or more efficient use of new data.