S. Sabesan, K. Narayanan, A. Prasad, A. Spanias, and L.D. Iasemidis (USA)
Information transfer, time series analysis, statistical and probabilistic modeling, nonlinear dynamics
A recently proposed measure, namely transfer entropy, is used to estimate the statistical dependence and direction of information flow between coupled systems. In this study, we suggest improvements in the selection of the parameters that significantly enhance the robustness of this measure in identifying the direction of information flow and quantifying the level of interaction between two observed data series. We demonstrate the potential usefulness of this method with simulation examples and show the statistical significance of the results using surrogate data.
Important Links:
Go Back
