Data-Driven modeling of Dynamical Systems: A Systems Theoretic Perspective
Limited Seating. No reservation required.
Abstract: In this talk, we will investigate various approaches to modeling dynamical systems from data. We will consider both frequency-domain and time-domain measurements of a dynamical system using systems theoretical concepts. In the former, data will correspond to the samples of a transfer function and we will show how to use these samples to learn reduced-order dynamics via rational interpolation and rational least-squares fitting. We will also extend these ideas to present a data-driven formulation for balanced truncation. In the case of time-domain data, we will assume access to (a subset of) state snapshots and use a least-squares minimization to learn the dynamics.
Bio: Serkan Gugercin is a professor of Mathematics at Virginia Tech. He holds the Class of 1950 Professorship. He is also a core faculty member and a Deputy Director in the Division of Computational Modeling and Data Analytics. In 1992, he received his B.S. degree in Electrical and Electronics Engineering from Middle East Technical University, Ankara, Turkey; and his M.S. and Ph.D. degrees in Electrical Engineering from Rice University, in 1999 and 2003, respectively. His primary research interests are model reduction, data-driven modeling, numerical linear algebra, approximation theory, and systems and control theory.
Dr. Gugercin received the Ralph Budd Award for Research in Engineering from Rice University in 2003 for the best doctoral thesis in the School of Engineering; Teaching Award from Jacobs University Bremen, in 2003; the National Science Foundation Early CAREER Award in Computational and Applied Mathematics in 2007; and the Alexander von Humboldt Research Fellowship in 2016.