Chapter

Mathematical Methods in Signal Processing and Digital Image Analysis

Part of the series Understanding Complex Systems pp 1-40

Multivariate Time Series Analysis

  • Björn SchelterAffiliated withFreiburg Center for Data Analysis and Modeling, University of Freiburg
  • , Rainer DahlhausAffiliated withInstitute for Applied Mathematics, University of Heidelberg
  • , Lutz LeistritzAffiliated withInstitute for Medical Statistics, Informatics, and Documentation, University of Jena
  • , Wolfram HesseAffiliated withInstitute for Medical Statistics, Informatics, and Documentation, University of Jena
  • , Bärbel SchackAffiliated withInstitute for Medical Statistics, Informatics, and Documentation, University of Jena
  • , Jürgen KurthsAffiliated withInstitute for Physics, University of Potsdam
  • , Jens TimmerAffiliated withFreiburg Center for Data Analysis and Modeling, University of Freiburg
  • , Herbert WitteAffiliated withInstitute for Medical Statistics, Informatics, and Documentation, University of Jena

* Final gross prices may vary according to local VAT.

Get Access

Abstract

Nowadays, modern measurement devices are capable to deliver signals with increasing data rates and higher spatial resolutions. When analyzing these data, particular interest is focused on disentangling the network structure underlying the recorded signals. Neither univariate nor bivariate analysis techniques are expected to describe the interactions between the processes sufficiently well. Moreover, the direction of the direct interactions is particularly important to understand the underlying network structure sufficiently well. Here, we present multivariate approaches to time series analysis being able to distinguish direct and indirect, in some cases the directions of interactions in linear as well as nonlinear systems.