Title: Meta-Learning and Automatic Machine Learning

Abstract

In the last decades, machine learning algorithms could impact vast plural areas of knowledge, creating  new challenges to democratize their usage. Selecting and tuning algorithms need to be performed by the vastest users' background. Facing these issues, Meta-learning, or learning to learn, is the science of systematically observing how different machine-learning approaches perform on a wide range of learning tasks and then learning from this experience or meta-data to learn new tasks much faster than otherwise possible. In recent years, Meta-learning has been increasing in importance and arose as an active field of research has developed around automated machine learning(AutoML).

Referente: Paolo Ceravolo