Deep learning has been studied for a few years for the modeling of complex physical processes in scientific and industrial fields such as environment, health, aeronautics, or energy production.  This area of research, although still emerging, is rapidly gaining momentum and is developing as an interdisciplinary field. It raises new challenges for the interaction between machine learning and physics. This tutorial will provide an introduction to physics-aware machine learning focusing on the modeling of dynamics characterizing physical phenomena in domains such as geophysics, climate, biology, fluid dynamics, etc. It will cover the main challenges in the field, and introduce the approaches developed so far ranging from pure ML methods to models that rely on a close integration of physical and ML components. The tutorial will address the specific generalization challenges raised by the modeling of physical systems and provide illustrations through different use cases.

It is intended for researchers and practitioners in the ML and Physics fields, who wish to have a better understanding of this exciting field and of recent research directions.