SKM 2023 – wissenschaftliches Programm
Bereiche | Tage | Auswahl | Suche | Aktualisierungen | Downloads | Hilfe
SOE: Fachverband Physik sozio-ökonomischer Systeme
SOE 13: Data Analytics of Complex Dynamical Systems (joint session DY/SOE)
SOE 13.8: Vortrag
Donnerstag, 30. März 2023, 11:45–12:00, MOL 213
Discovering Causality and Coupling in high dimensional nonlinear dynamical systems — •Tim Kroll1,2 and Oliver Kamps2 — 1Westfälische Wilhelms-Universität Münster, 48149 Münster — 2Center for Nonlinear Science, Westfälische Wilhelms-Universität Münster, 48149 Münster
In this talk we present a method to infer causal relationships between observables from data of systems where the underlying dynamics are not known a priori. The method is based on the hypothesis that the system of interest can in principle be described by a set of coupled nonlinear ordinary differential equations. Following the work of Prusseit and Lehnertz in 2008 we can then determine the couplings between observables by integrating out all other observables. Since the estimation of the underlying dynamical system invokes the efficient representation in terms of polynomials the method can be applied also to hight dimensional systems. We demonstrate the capabilities of the method by inferring the network structure of coupled Rössler-Attractors.