Overview and Unifying Conceptualization of Automated Machine Learning

Abstract

We introduce a novel generic mathematical formulation of AutoML, resting on formal definitions of hyperparameter optimization (HPO) and meta-learning. In light of this formulation, we decompose various algorithms and show that HPO does not really address the AutoML problem more than classical machine learning algorithms, while meta-learning does.

Other branches such as transfer learning and ensemble learning are also reviewed, re-formulated and unified, providing a formal algebraic language to facilitate future research.