scientific article; zbMATH DE number 2169745

From MaRDI portal
Publication:4677204





Let \(\Xi=\{\xi_{t}\), \(t\in T\}\) be a family of jointly Gaussian random variables, \(E\xi_{t}=0\). The space \(SG_{\Xi} (\Omega)\) is called the space of square-Gaussian random variables with respect to \(\Xi\), if the random variables from \(SG_{\Xi} (\Omega)\) can be presented in the form \(\zeta=\overline\xi^{T}A\overline\xi-E\overline\xi^{T}A\overline\xi\), where \(\overline\xi={\{\xi_1,\ldots,\xi_{d}\}}^{T}\) is a Gaussian random vector, \(E\overline\xi=0\), the random variables \(\xi_{i}\), \(i=1,\ldots,d\), are from \(\Xi\), \(A\) is a symmetric matrix, or the random variables from \(SG_{\Xi} (\Omega)\) are mean-square limits of sequences of random variables that can be presented in the above mentioned form. A random process \(\{\zeta(t)\), \(t\in T\}\) is called square-Gaussian random process if the random variable \(\zeta(t)\) for all \(t\) belongs to the space \(SG_{\Xi} (\Omega)\) and \(sup_{t\in T}E\zeta^2(t)<\infty\). Using inequalities for distributions of quadratic forms of square-Gaussian random variables and their mean-square limits, the author constructs a criterion for testing hypothesis about the covariance function of Gaussian stationary random processes and homogeneous and isotropic Gaussian random fields for given estimations of covariance functions.











This page was built for publication:

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4677204)