|
Short Communications
Exact lower and upper bounds for Gaussian measures
I. Pinelis Department of Mathematical Sciences, Michigan Technological University, Houghton, MI, USA
Abstract:
Exact upper and lower bounds on the ratio
$\operatorname{\mathbf{E}}w(\mathbf{X}-\mathbf{v})/\operatorname{\mathbf{E}}w(\mathbf{X})$ for a centered
Gaussian random vector $\mathbf{X}$ in $\mathbf{R}^n$ are obtained, as well as bounds on
the rate of change of $\operatorname{\mathbf{E}}w(\mathbf{X}-t\mathbf{v})$ in $t$, where
$w\colon\mathbf{R}^n\to[0,\infty)$ is any even unimodal function and
$\mathbf{v}$ is any vector in $\mathbf{R}^n$. As a corollary of such results,
exact upper and lower bounds on the power function of statistical tests for
the mean of a multivariate normal distribution are given.
Keywords:
Gaussian measures, multivariate normal distribution, shifts, unimodality, logconcavity, monotonicity, exact bounds, tests for the mean.
Received: 18.04.2019
Published: 22.07.2022
Citation:
I. Pinelis, “Exact lower and upper bounds for Gaussian measures”, Teor. Veroyatnost. i Primenen., 67:3 (2022), 607–617; Theory Probab. Appl., 67:3 (2022), 485–493
Linking options:
https://www.mathnet.ru/eng/tvp5316https://doi.org/10.4213/tvp5316 https://www.mathnet.ru/eng/tvp/v67/i3/p607
|
| Statistics & downloads: |
| Abstract page: | 266 | | Full-text PDF : | 86 | | References: | 107 | | First page: | 10 |
|