Correlation- and Causality Networks of Complex Systems

The research objective is to reconstruct interactions between components of a complex system from recorded time series. The time series result from sequential measuring of interesting state variables as, for instance, concentrations of biological/chemical species or nervous cell activity. Correlations express statistical relations between components, however, per se they do not evince a causal link, i.e. "who causes whom". Beyond the detection of such causal connections via appropriate mathematical approaches (predictive information flow, Granger causality) the topological structure is represented via network graphs.