Главная

Daniel Gervini

  1. Warped Functional Regression.

    Пнд, 03/12/2012 - 15:01 — SomNambul
    Authors: Daniel Gervini
    Subjects: Methodology
    Abstract

    A characteristic feature of functional data is the presence of time
    variability in addition to amplitude variability. The existing functional
    regression methods do not handle time variability in an explicit and efficient
    way. In this paper we introduce a functional regression method that
    incorporates time warping as an intrinsic part of the model. The method
    achieves good predictive power in a parsimonious way, and allows for unified
    statistical inference of time and amplitude variability.

  2. Dynamic Functional Regression.

    Втр, 03/15/2011 - 16:01 — SomNambul
    Authors: Daniel Gervini
    Subjects: Methodology
    Abstract

    A characteristic feature of samples of curves is the presence of time
    variability in addition to amplitude variability. The existing functional
    regression methods do not handle time variability in an efficient manner. We
    propose in this paper a regression method that incorporates time warping as an
    intrinsic part of the model. In this way, the method attains a high predictive
    power in a parsimonious and efficient manner, avoiding overfitting and
    simplifying statistical inference.

  3. Detecting and handling outlying trajectories in irregularly sampled functional datasets.

    Ср, 11/03/2010 - 16:01 — SomNambul
    Authors: Daniel Gervini
    Subjects: Applications
    Abstract

    Outlying curves often occur in functional or longitudinal datasets, and can
    be very influential on parameter estimators and very hard to detect visually.
    In this article we introduce estimators of the mean and the principal
    components that are resistant to, and then can be used for detection of,
    outlying sample trajectories.

  4. Outlier detection and trimmed estimation in general functional spaces.

    Пт, 01/08/2010 - 16:01 — SomNambul
    Authors: Daniel Gervini
    Subjects: Methodology
    Abstract

    This article introduces trimmed estimators for the mean and covariance
    functional of data in general Hilbert spaces. The estimators are based on a
    data depth measure that can be computed on any Hilbert space, because it is
    defined only in terms of the interdistances between data points. We show that
    the estimators can attain the maximum breakdown point by properly choosing the
    tuning parameters, and that they possess better outlier resistance properties
    than alternative estimators, as shown by a comparative Monte Carlo study.

Math-arch

RSS-материал