This lecture was given by Prof. Peter J. Schmid, Imperial College London, UK in the framework of the von Karman Lecture Series on Machine Learning for Fluid Mechanics organized by the von Karman Institute and the Université libre de Bruxelles in February 2020.
Dynamic mode decomposition is a simple, yet effective data-driven tool to extract quantitative information about fluid flows from sequences of measurements. In its simplest form, it is straightforward to implement and use. Nonetheless, it is based on a sound, sophisticated and rich mathematical formalism - Koopman analysis - that provides guidance and support for further developments, applications and interpretations. Many extensions and improvements of the original algorithm are available, and a wide range of applications (to numerical and experimental data) can be found in a growing body of literature.