Rectangular regions of maximum probability content
The paper proves the following theorem. Let X be a bivariate random vector with pdf of the form \(f(x)=g((x-\mu)\Sigma^{-1}(x-\mu)')\), where \(\mu\in R^ 2\), \(\Sigma^{-1}=(\sigma^{ij})\) is a 2\(\times 2\) symmetric positive definite matrix, and the function g:[0,\(\infty)\to (0,\infty)\) is strictly increasing. Then for any fixed \(\lambda >0\) the set with maximal probability content in the class \(R_{\lambda}=\{R:\) R is a rectangle in \(R^ 2\) with sides parallel to the axes and \(\int_{R}dx=\lambda\}\) is \(\{(x_ 1,x_ 2):| x_ 1-\mu_ 1| <c\), \(| x_ 2-\mu_ 2| <\lambda /4c\}\), with \(c=[(\lambda /4)(\sigma^{22}/\sigma^{11})^{1/2}]^{1/2}\).
- scientific article; zbMATH DE number 3898022
- scientific article; zbMATH DE number 562226
- Derivation of bivariate probability density functions with exponential marginals
- Confidence regions for bivariate probability density functions using polygonal areas
- Region of maximum probability content of fixed diameter
This page was built for publication: Rectangular regions of maximum probability content
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q799019)
