Semiparametric non-linear time series model selection

Jiti Gao, H. Tong

    Research output: Contribution to journalArticle

    26 Citations (Scopus)

    Abstract

    Semiparametric time series regression is often used without checking its suitability,resulting in an unnecessarily complicated model. In practice, one may encounter computationaldifficulties caused by the curse of dimensionality.The paper suggests that to provide more precisepredictions we need to choose the most significant regressors for both the parametric andthe nonparametric time series components.We develop a novel cross-validation-based modelselection procedure for the simultaneous choice of both the parametric and the nonparametrictime series components, and we establish some asymptotic properties of the model selectionprocedure proposed. In addition, we demonstrate how to implement it by using both simulatedand real examples. Our empirical studies show that the procedure works well.
    Original languageEnglish
    Pages (from-to)321-336
    JournalJournal of the Royal Statistical Society Series B
    Volume66
    Issue number2
    DOIs
    Publication statusPublished - 2004

    Fingerprint

    Nonlinear Time Series Model
    Model Selection
    Time series
    Curse of Dimensionality
    Cross-validation
    Empirical Study
    Asymptotic Properties
    Choose
    Regression
    Series
    Model
    Demonstrate
    Time series models
    Nonlinear time series
    Model selection
    Empirical study
    Curse of dimensionality
    Asymptotic properties

    Cite this

    @article{b5a2fd980a2d4743812e6d4308669156,
    title = "Semiparametric non-linear time series model selection",
    abstract = "Semiparametric time series regression is often used without checking its suitability,resulting in an unnecessarily complicated model. In practice, one may encounter computationaldifficulties caused by the curse of dimensionality.The paper suggests that to provide more precisepredictions we need to choose the most significant regressors for both the parametric andthe nonparametric time series components.We develop a novel cross-validation-based modelselection procedure for the simultaneous choice of both the parametric and the nonparametrictime series components, and we establish some asymptotic properties of the model selectionprocedure proposed. In addition, we demonstrate how to implement it by using both simulatedand real examples. Our empirical studies show that the procedure works well.",
    author = "Jiti Gao and H. Tong",
    year = "2004",
    doi = "10.1111/j.1369-7412.2004.05303.x",
    language = "English",
    volume = "66",
    pages = "321--336",
    journal = "Journal of the Royal Statistical Society Series B",
    issn = "1369-7412",
    publisher = "Wiley-Blackwell",
    number = "2",

    }

    Semiparametric non-linear time series model selection. / Gao, Jiti; Tong, H.

    In: Journal of the Royal Statistical Society Series B, Vol. 66, No. 2, 2004, p. 321-336.

    Research output: Contribution to journalArticle

    TY - JOUR

    T1 - Semiparametric non-linear time series model selection

    AU - Gao, Jiti

    AU - Tong, H.

    PY - 2004

    Y1 - 2004

    N2 - Semiparametric time series regression is often used without checking its suitability,resulting in an unnecessarily complicated model. In practice, one may encounter computationaldifficulties caused by the curse of dimensionality.The paper suggests that to provide more precisepredictions we need to choose the most significant regressors for both the parametric andthe nonparametric time series components.We develop a novel cross-validation-based modelselection procedure for the simultaneous choice of both the parametric and the nonparametrictime series components, and we establish some asymptotic properties of the model selectionprocedure proposed. In addition, we demonstrate how to implement it by using both simulatedand real examples. Our empirical studies show that the procedure works well.

    AB - Semiparametric time series regression is often used without checking its suitability,resulting in an unnecessarily complicated model. In practice, one may encounter computationaldifficulties caused by the curse of dimensionality.The paper suggests that to provide more precisepredictions we need to choose the most significant regressors for both the parametric andthe nonparametric time series components.We develop a novel cross-validation-based modelselection procedure for the simultaneous choice of both the parametric and the nonparametrictime series components, and we establish some asymptotic properties of the model selectionprocedure proposed. In addition, we demonstrate how to implement it by using both simulatedand real examples. Our empirical studies show that the procedure works well.

    U2 - 10.1111/j.1369-7412.2004.05303.x

    DO - 10.1111/j.1369-7412.2004.05303.x

    M3 - Article

    VL - 66

    SP - 321

    EP - 336

    JO - Journal of the Royal Statistical Society Series B

    JF - Journal of the Royal Statistical Society Series B

    SN - 1369-7412

    IS - 2

    ER -