The goal of this workshop is to provide an overview of data-driven methods for the analysis of time-series data. Given only measurement or simulation data, these methods can be used to extract global properties of the underlying system such as time scales and metastable sets or to directly learn the governing equations of the system. Other topics of interest include multiscale modeling, homogenization, dimensionality reduction, manifold learning, and control as well as applications in molecular dynamics, fluid dynamics, climate science, and engineering.
The workshop is part of the Thematic Einstein Semester “The Mathematics of Complex Social Systems: Past, Present, and Future”.
The workshop takes place as a hybrid event, where the on-site part will be at Zuse Institute Berlin in Berlin-Dahlem.
|Stability analysis in Koopman operator theory: A data-driven approach
|The Koopman operator-based data-driven algorithms on nonautonomous and stochastic dynamics
|Efficient data-driven prediction and control of complex systems via the Koopman operator
|Koopman modes of the sea surface temperature in the Tropical Pacific
|Hybrid modeling for the stochastic simulation of spatial and non-spatial multi-scale chemical kinetics
|Collective variables and tipping analysis of agent-based models
|Understanding network dynamics with graph representation learning
|Machine learning of self organization from observation
|Manifold Learning 2.0: Explanations and eigenflows