On the asymptotic distribution of weighted uniform empirical and quantile processes in the middle and on the tails
Let [alpha]n and un be the uniform empirical and quantile processes. We investigate the asymptotic distribution of the suprema of [alpha]n(s)/(s(1 - s))1/2±[nu] and un(s)/(s(1 - s))1/2±[nu] with , when the supremum is taken over ranges, depending on n, in the middle of the interval [0, 1], near 0 and near 1. We show that with suitable norming factors the said asymptotic distributions can be radically different or the same, depending on the sign and value of [nu] in the weight function 1/(s(1 - s))1/2±[nu].
Year of publication: |
1985
|
---|---|
Authors: | Csörgo, Miklós ; Mason, David M. |
Published in: |
Stochastic Processes and their Applications. - Elsevier, ISSN 0304-4149. - Vol. 21.1985, 1, p. 119-132
|
Publisher: |
Elsevier |
Keywords: | weighted empirical and quantile processes asymptotic distributions approximations |
Saved in:
Saved in favorites
Similar items by person
-
Moderate deviations and intermediate efficiency for lack-of-fit tests
Mason, David M., (2012)
-
Matrix normalized convergence of a Lévy process to normality at zero
Maller, Ross A., (2015)
-
Randomly weighted self-normalized Lévy processes
Kevei, Péter, (2013)
- More ...