Robust model selection for a semimartingale continuous time regression from discrete data

Konev Victor, Pergamenchtchikov Serguei

    Research output: Contribution to journalArticle

    5 Citations (Scopus)

    Abstract

    The paper considers the problem of estimating a periodic function in a continuous time regression model observed under a general semimartingale noise with an unknown distribution in the case when continuous observation cannot be provided and only discrete time measurements are available. Two specific types of noises are studied in detail: a non-Gaussian Ornstein-Uhlenbeck process and a time-varying linear combination of a Brownian motion and compound Poisson process. We develop new analytical tools to treat the adaptive estimation problems from discrete data. A lower bound for the frequency sampling, needed for the efficiency of the procedure constructed by discrete observations, has been found. Sharp non-asymptotic oracle inequalities for the robust quadratic risk have been derived. New convergence rates for the efficient procedures have been obtained. An example of the regression with a martingale noise exhibits that the minimax robust convergence rate may be both higher or lower as compared with the minimax rate for the "white noise" model. The results of Monte-Carlo simulations are given.

    Original languageEnglish
    Pages (from-to)294-326
    Number of pages33
    JournalStochastic Processes and their Applications
    Volume125
    Issue number1
    DOIs
    Publication statusPublished - Jan 2015

    Fingerprint

    Discrete Data
    Semimartingale
    Model Selection
    Convergence Rate
    Continuous Time
    Regression
    White Noise Model
    Oracle Inequalities
    Discrete Observations
    Minimax Rate
    Compound Poisson Process
    Adaptive Estimation
    Continuous-time Model
    Ornstein-Uhlenbeck Process
    Brownian movement
    White noise
    Periodic Functions
    Time measurement
    Minimax
    Martingale

    Keywords

    • Estimation from discrete data
    • Model selection
    • Robust risk estimation
    • Semimartingale regression
    • Sharp oracle inequalities

    ASJC Scopus subject areas

    • Modelling and Simulation
    • Statistics and Probability
    • Applied Mathematics

    Cite this

    Robust model selection for a semimartingale continuous time regression from discrete data. / Victor, Konev; Serguei, Pergamenchtchikov.

    In: Stochastic Processes and their Applications, Vol. 125, No. 1, 01.2015, p. 294-326.

    Research output: Contribution to journalArticle

    Victor, Konev ; Serguei, Pergamenchtchikov. / Robust model selection for a semimartingale continuous time regression from discrete data. In: Stochastic Processes and their Applications. 2015 ; Vol. 125, No. 1. pp. 294-326.
    @article{74d86cc98ee446309eccebc6dbd9bc5b,
    title = "Robust model selection for a semimartingale continuous time regression from discrete data",
    abstract = "The paper considers the problem of estimating a periodic function in a continuous time regression model observed under a general semimartingale noise with an unknown distribution in the case when continuous observation cannot be provided and only discrete time measurements are available. Two specific types of noises are studied in detail: a non-Gaussian Ornstein-Uhlenbeck process and a time-varying linear combination of a Brownian motion and compound Poisson process. We develop new analytical tools to treat the adaptive estimation problems from discrete data. A lower bound for the frequency sampling, needed for the efficiency of the procedure constructed by discrete observations, has been found. Sharp non-asymptotic oracle inequalities for the robust quadratic risk have been derived. New convergence rates for the efficient procedures have been obtained. An example of the regression with a martingale noise exhibits that the minimax robust convergence rate may be both higher or lower as compared with the minimax rate for the {"}white noise{"} model. The results of Monte-Carlo simulations are given.",
    keywords = "Estimation from discrete data, Model selection, Robust risk estimation, Semimartingale regression, Sharp oracle inequalities",
    author = "Konev Victor and Pergamenchtchikov Serguei",
    year = "2015",
    month = "1",
    doi = "10.1016/j.spa.2014.08.003",
    language = "English",
    volume = "125",
    pages = "294--326",
    journal = "Stochastic Processes and their Applications",
    issn = "0304-4149",
    publisher = "Elsevier",
    number = "1",

    }

    TY - JOUR

    T1 - Robust model selection for a semimartingale continuous time regression from discrete data

    AU - Victor, Konev

    AU - Serguei, Pergamenchtchikov

    PY - 2015/1

    Y1 - 2015/1

    N2 - The paper considers the problem of estimating a periodic function in a continuous time regression model observed under a general semimartingale noise with an unknown distribution in the case when continuous observation cannot be provided and only discrete time measurements are available. Two specific types of noises are studied in detail: a non-Gaussian Ornstein-Uhlenbeck process and a time-varying linear combination of a Brownian motion and compound Poisson process. We develop new analytical tools to treat the adaptive estimation problems from discrete data. A lower bound for the frequency sampling, needed for the efficiency of the procedure constructed by discrete observations, has been found. Sharp non-asymptotic oracle inequalities for the robust quadratic risk have been derived. New convergence rates for the efficient procedures have been obtained. An example of the regression with a martingale noise exhibits that the minimax robust convergence rate may be both higher or lower as compared with the minimax rate for the "white noise" model. The results of Monte-Carlo simulations are given.

    AB - The paper considers the problem of estimating a periodic function in a continuous time regression model observed under a general semimartingale noise with an unknown distribution in the case when continuous observation cannot be provided and only discrete time measurements are available. Two specific types of noises are studied in detail: a non-Gaussian Ornstein-Uhlenbeck process and a time-varying linear combination of a Brownian motion and compound Poisson process. We develop new analytical tools to treat the adaptive estimation problems from discrete data. A lower bound for the frequency sampling, needed for the efficiency of the procedure constructed by discrete observations, has been found. Sharp non-asymptotic oracle inequalities for the robust quadratic risk have been derived. New convergence rates for the efficient procedures have been obtained. An example of the regression with a martingale noise exhibits that the minimax robust convergence rate may be both higher or lower as compared with the minimax rate for the "white noise" model. The results of Monte-Carlo simulations are given.

    KW - Estimation from discrete data

    KW - Model selection

    KW - Robust risk estimation

    KW - Semimartingale regression

    KW - Sharp oracle inequalities

    UR - http://www.scopus.com/inward/record.url?scp=84907695590&partnerID=8YFLogxK

    UR - http://www.scopus.com/inward/citedby.url?scp=84907695590&partnerID=8YFLogxK

    U2 - 10.1016/j.spa.2014.08.003

    DO - 10.1016/j.spa.2014.08.003

    M3 - Article

    VL - 125

    SP - 294

    EP - 326

    JO - Stochastic Processes and their Applications

    JF - Stochastic Processes and their Applications

    SN - 0304-4149

    IS - 1

    ER -