Maximum likelihood estimation of parameters under a spatial sampling scheme (Q1314469)

From MaRDI portal
scientific article
Language Label Description Also known as
English
Maximum likelihood estimation of parameters under a spatial sampling scheme
scientific article

    Statements

    Maximum likelihood estimation of parameters under a spatial sampling scheme (English)
    0 references
    0 references
    25 October 1994
    0 references
    Let \(X(t)\), \(t=(t_ 1, \dots, t_ d) \in [0,1]^ d\), be a spatial Gaussian process with a multiplicative Ornstein-Uhlenbeck covariance function \[ \Gamma (t,s) = \sigma^ 2 \exp \left\{ - \sum^ d_{j=1} \theta_ j | t_ j-s_ j | \right\} \] where \(\theta_ j\) \((j=1,\dots,d)\) and \(\sigma^ 2\) are unknown parameters. In some statistical problems, it is necessary to estimate these parameters. Let \(\widehat \theta_ j\) \((j = 1,\dots,d)\) and \(\widehat \sigma^ 2\) be maximum likelihood estimators of the parameters \(\theta_ j\) \((j=1,\dots,d)\) and \(\sigma^ 2\), respectively, based on a set of observations which consists of \[ X \biggl( \bigl( u_{k_ 1}^{(1)}, \dots, u_{k_ d}^{(d)} \bigr) \biggr);\;1 \leq k_ i \leq n_ i,\;i=1,\dots,d. \] The author proves consistency and asymptotic normality of these maximum likelihood estimators under some conditions.
    0 references
    Gaussian random fields
    0 references
    lattice sampling
    0 references
    computer experiments
    0 references
    spatial Gaussian process
    0 references
    multiplicative Ornstein-Uhlenbeck covariance function
    0 references
    maximum likelihood estimators
    0 references
    consistency
    0 references
    asymptotic normality
    0 references

    Identifiers

    0 references
    0 references
    0 references
    0 references
    0 references