Causal inference in nonlinear systems

Granger causality versus time-delayed mutual information

Songting Li, Yanyang Xiao, Doug Zhou, David Cai

    Research output: Contribution to journalArticle

    Abstract

    The Granger causality (GC) analysis has been extensively applied to infer causal interactions in dynamical systems arising from economy and finance, physics, bioinformatics, neuroscience, social science, and many other fields. In the presence of potential nonlinearity in these systems, the validity of the GC analysis in general is questionable. To illustrate this, here we first construct minimal nonlinear systems and show that the GC analysis fails to infer causal relations in these systems - it gives rise to all types of incorrect causal directions. In contrast, we show that the time-delayed mutual information (TDMI) analysis is able to successfully identify the direction of interactions underlying these nonlinear systems. We then apply both methods to neuroscience data collected from experiments and demonstrate that the TDMI analysis but not the GC analysis can identify the direction of interactions among neuronal signals. Our work exemplifies inference hazards in the GC analysis in nonlinear systems and suggests that the TDMI analysis can be an appropriate tool in such a case.

    Original languageEnglish (US)
    Article number052216
    JournalPhysical Review E
    Volume97
    Issue number5
    DOIs
    StatePublished - May 29 2018

    Fingerprint

    information analysis
    Causal Inference
    Granger Causality
    nonlinear systems
    Mutual Information
    inference
    neurology
    Nonlinear Systems
    finance
    interactions
    economy
    dynamical systems
    hazards
    Neuroscience
    nonlinearity
    physics
    Interaction
    Social Sciences
    Finance
    Hazard

    ASJC Scopus subject areas

    • Statistical and Nonlinear Physics
    • Statistics and Probability
    • Condensed Matter Physics

    Cite this

    Causal inference in nonlinear systems : Granger causality versus time-delayed mutual information. / Li, Songting; Xiao, Yanyang; Zhou, Doug; Cai, David.

    In: Physical Review E, Vol. 97, No. 5, 052216, 29.05.2018.

    Research output: Contribution to journalArticle

    Li, Songting ; Xiao, Yanyang ; Zhou, Doug ; Cai, David. / Causal inference in nonlinear systems : Granger causality versus time-delayed mutual information. In: Physical Review E. 2018 ; Vol. 97, No. 5.
    @article{154eb24f721c4f20af4fcd38dba362b9,
    title = "Causal inference in nonlinear systems: Granger causality versus time-delayed mutual information",
    abstract = "The Granger causality (GC) analysis has been extensively applied to infer causal interactions in dynamical systems arising from economy and finance, physics, bioinformatics, neuroscience, social science, and many other fields. In the presence of potential nonlinearity in these systems, the validity of the GC analysis in general is questionable. To illustrate this, here we first construct minimal nonlinear systems and show that the GC analysis fails to infer causal relations in these systems - it gives rise to all types of incorrect causal directions. In contrast, we show that the time-delayed mutual information (TDMI) analysis is able to successfully identify the direction of interactions underlying these nonlinear systems. We then apply both methods to neuroscience data collected from experiments and demonstrate that the TDMI analysis but not the GC analysis can identify the direction of interactions among neuronal signals. Our work exemplifies inference hazards in the GC analysis in nonlinear systems and suggests that the TDMI analysis can be an appropriate tool in such a case.",
    author = "Songting Li and Yanyang Xiao and Doug Zhou and David Cai",
    year = "2018",
    month = "5",
    day = "29",
    doi = "10.1103/PhysRevE.97.052216",
    language = "English (US)",
    volume = "97",
    journal = "Physical review. E",
    issn = "1539-3755",
    publisher = "American Physical Society",
    number = "5",

    }

    TY - JOUR

    T1 - Causal inference in nonlinear systems

    T2 - Granger causality versus time-delayed mutual information

    AU - Li, Songting

    AU - Xiao, Yanyang

    AU - Zhou, Doug

    AU - Cai, David

    PY - 2018/5/29

    Y1 - 2018/5/29

    N2 - The Granger causality (GC) analysis has been extensively applied to infer causal interactions in dynamical systems arising from economy and finance, physics, bioinformatics, neuroscience, social science, and many other fields. In the presence of potential nonlinearity in these systems, the validity of the GC analysis in general is questionable. To illustrate this, here we first construct minimal nonlinear systems and show that the GC analysis fails to infer causal relations in these systems - it gives rise to all types of incorrect causal directions. In contrast, we show that the time-delayed mutual information (TDMI) analysis is able to successfully identify the direction of interactions underlying these nonlinear systems. We then apply both methods to neuroscience data collected from experiments and demonstrate that the TDMI analysis but not the GC analysis can identify the direction of interactions among neuronal signals. Our work exemplifies inference hazards in the GC analysis in nonlinear systems and suggests that the TDMI analysis can be an appropriate tool in such a case.

    AB - The Granger causality (GC) analysis has been extensively applied to infer causal interactions in dynamical systems arising from economy and finance, physics, bioinformatics, neuroscience, social science, and many other fields. In the presence of potential nonlinearity in these systems, the validity of the GC analysis in general is questionable. To illustrate this, here we first construct minimal nonlinear systems and show that the GC analysis fails to infer causal relations in these systems - it gives rise to all types of incorrect causal directions. In contrast, we show that the time-delayed mutual information (TDMI) analysis is able to successfully identify the direction of interactions underlying these nonlinear systems. We then apply both methods to neuroscience data collected from experiments and demonstrate that the TDMI analysis but not the GC analysis can identify the direction of interactions among neuronal signals. Our work exemplifies inference hazards in the GC analysis in nonlinear systems and suggests that the TDMI analysis can be an appropriate tool in such a case.

    UR - http://www.scopus.com/inward/record.url?scp=85047758273&partnerID=8YFLogxK

    UR - http://www.scopus.com/inward/citedby.url?scp=85047758273&partnerID=8YFLogxK

    U2 - 10.1103/PhysRevE.97.052216

    DO - 10.1103/PhysRevE.97.052216

    M3 - Article

    VL - 97

    JO - Physical review. E

    JF - Physical review. E

    SN - 1539-3755

    IS - 5

    M1 - 052216

    ER -