A Note on Dispersive Ordering of Record Values
A sequence of i.i.d. random values \((X_n)^\infty_1\) with the same absolutely continuous distribution function \(F(x)\) \((F(0)=0)\) is considered. The corresponding sequence of its record values \((R_n)^\infty_1\) is investigated. The sequence of records is proved not to decrease in dispersion if \(F\) determines a decreasing hazard rate. The author defines decreasing in dispersion according to the following relation for random values \(X\) and \(Y\) with distribution functions \(F_X\) and \(F_Y :X\) is less dispersed than \(Y\) if \[ (\forall \alpha,\beta: 0\leq\alpha <\beta\leq 1)\;F_X^{-1} (\beta)-F_X^{-1} (\alpha) \leq F^{-1}_Y(\beta) -F_Y^{-1}(\alpha). \] Some other equivalent definitions of this ordering are discussed. An example of the function \(F\), when the corresponding sequence of records is not non-decreasing in dispersion, is offered.
- On stochastic orderings between residual record values
- STOCHASTIC ORDERINGS OF DISCRETE-TIME PROCESSES AND DISCRETE RECORD VALUES
- Stochastic ordering of random kth record values
- The extropy of order statistics and record values
- On characterizations based on record values and order statistics
- scientific article; zbMATH DE number 1014852
- On the comparison of relative spacings with applications
- On stochastic orderings between residual record values
- Runs based on records: their distributional properties and an application to testing for dispersive ordering
- Stochastic Ordering of the Number of Records
- Some stochastic comparisons of lower records and lower record spacings
- Dispersive ordering-some applications and examples
- A new dependence ordering with applications
This page was built for publication: A Note on Dispersive Ordering of Record Values
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q4215892)
