Comparison of Granger Causality and Phase Slope Index

Guido Nolte, Andreas Ziehe, Nicole Krämer, Florin Popescu, Klaus-Robert Müller
Proceedings of Workshop on Causality: Objectives and Assessment at NIPS 2008, PMLR 6:267-276, 2010.

Abstract

We recently proposed a new measure, termed Phase Slope Index (PSI), It estimates the causal direction of interactions robustly with respect to instantaneous mixtures of independent sources with arbitrary spectral content. We compared this method to Granger Causality for linear systems containing spatially and temporarily mixed noise and found that, in contrast to PSI, the latter was not able to properly distinguish truly interacting systems from mixed noise. Here, we extent this analysis with respect to two aspects: a) we analyze Granger causality and PSI also for non-mixed noise, and b) we analyze PSI for nonlinear interactions. We found a) that Granger causality, in contrast to PSI, fails also for non-mixed noise if the memory-time of the sender of information is long compared to the transmission time of the information, and b) that PSI, being a linear method, eventually misses nonlinear interactions but is unlikely to give false positive results.

Cite this Paper

BibTeX
@InProceedings{pmlr-v6-nolte10a, title = {Comparison of Granger Causality and Phase Slope Index}, author = {Nolte, Guido and Ziehe, Andreas and Krämer, Nicole and Popescu, Florin and Müller, Klaus-Robert}, booktitle = {Proceedings of Workshop on Causality: Objectives and Assessment at NIPS 2008}, pages = {267--276}, year = {2010}, editor = {Guyon, Isabelle and Janzing, Dominik and Schölkopf, Bernhard}, volume = {6}, series = {Proceedings of Machine Learning Research}, address = {Whistler, Canada}, month = {12 Dec}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v6/nolte10a/nolte10a.pdf}, url = {https://proceedings.mlr.press/v6/nolte10a.html}, abstract = {We recently proposed a new measure, termed Phase Slope Index (PSI), It estimates the causal direction of interactions robustly with respect to instantaneous mixtures of independent sources with arbitrary spectral content. We compared this method to Granger Causality for linear systems containing spatially and temporarily mixed noise and found that, in contrast to PSI, the latter was not able to properly distinguish truly interacting systems from mixed noise. Here, we extent this analysis with respect to two aspects: a) we analyze Granger causality and PSI also for non-mixed noise, and b) we analyze PSI for nonlinear interactions. We found a) that Granger causality, in contrast to PSI, fails also for non-mixed noise if the memory-time of the sender of information is long compared to the transmission time of the information, and b) that PSI, being a linear method, eventually misses nonlinear interactions but is unlikely to give false positive results.} }
Endnote
%0 Conference Paper %T Comparison of Granger Causality and Phase Slope Index %A Guido Nolte %A Andreas Ziehe %A Nicole Krämer %A Florin Popescu %A Klaus-Robert Müller %B Proceedings of Workshop on Causality: Objectives and Assessment at NIPS 2008 %C Proceedings of Machine Learning Research %D 2010 %E Isabelle Guyon %E Dominik Janzing %E Bernhard Schölkopf %F pmlr-v6-nolte10a %I PMLR %P 267--276 %U https://proceedings.mlr.press/v6/nolte10a.html %V 6 %X We recently proposed a new measure, termed Phase Slope Index (PSI), It estimates the causal direction of interactions robustly with respect to instantaneous mixtures of independent sources with arbitrary spectral content. We compared this method to Granger Causality for linear systems containing spatially and temporarily mixed noise and found that, in contrast to PSI, the latter was not able to properly distinguish truly interacting systems from mixed noise. Here, we extent this analysis with respect to two aspects: a) we analyze Granger causality and PSI also for non-mixed noise, and b) we analyze PSI for nonlinear interactions. We found a) that Granger causality, in contrast to PSI, fails also for non-mixed noise if the memory-time of the sender of information is long compared to the transmission time of the information, and b) that PSI, being a linear method, eventually misses nonlinear interactions but is unlikely to give false positive results.
RIS
TY - CPAPER TI - Comparison of Granger Causality and Phase Slope Index AU - Guido Nolte AU - Andreas Ziehe AU - Nicole Krämer AU - Florin Popescu AU - Klaus-Robert Müller BT - Proceedings of Workshop on Causality: Objectives and Assessment at NIPS 2008 DA - 2010/02/18 ED - Isabelle Guyon ED - Dominik Janzing ED - Bernhard Schölkopf ID - pmlr-v6-nolte10a PB - PMLR DP - Proceedings of Machine Learning Research VL - 6 SP - 267 EP - 276 L1 - http://proceedings.mlr.press/v6/nolte10a/nolte10a.pdf UR - https://proceedings.mlr.press/v6/nolte10a.html AB - We recently proposed a new measure, termed Phase Slope Index (PSI), It estimates the causal direction of interactions robustly with respect to instantaneous mixtures of independent sources with arbitrary spectral content. We compared this method to Granger Causality for linear systems containing spatially and temporarily mixed noise and found that, in contrast to PSI, the latter was not able to properly distinguish truly interacting systems from mixed noise. Here, we extent this analysis with respect to two aspects: a) we analyze Granger causality and PSI also for non-mixed noise, and b) we analyze PSI for nonlinear interactions. We found a) that Granger causality, in contrast to PSI, fails also for non-mixed noise if the memory-time of the sender of information is long compared to the transmission time of the information, and b) that PSI, being a linear method, eventually misses nonlinear interactions but is unlikely to give false positive results. ER -
APA
Nolte, G., Ziehe, A., Krämer, N., Popescu, F. & Müller, K.. (2010). Comparison of Granger Causality and Phase Slope Index. Proceedings of Workshop on Causality: Objectives and Assessment at NIPS 2008, in Proceedings of Machine Learning Research 6:267-276 Available from https://proceedings.mlr.press/v6/nolte10a.html.

Related Material