Extremes of a random number of variables from periodic sequences (Q1890872)

From MaRDI portal
Revision as of 21:52, 28 July 2023 by Importer (talk | contribs) (‎Created a new Item)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
scientific article
Language Label Description Also known as
English
Extremes of a random number of variables from periodic sequences
scientific article

    Statements

    Extremes of a random number of variables from periodic sequences (English)
    0 references
    0 references
    18 September 1995
    0 references
    Let \(\{X_ n\}\) be a periodic sequence of random variables, i.e. there exists \(T \in \mathbb{N}\) such that for every choice of integers \(i_ 1, \dots, i_ n\) the random vectors \((X_{i_ 1}, \dots, X_{i_ n})\) and \((X_{i_ 1 + T}, \dots, X_{i_ n + T})\) are identically distributed. If \(T\) is minimal, then the sequence is called a \(T\)- periodic sequence. The author considers the asymptotic behaviour of \(S_ m [X,u_ n] (B) = \sum^ m_{i = 1} \text{\textbf{1}}_{\{X_ i > u_ n\}} \delta_{i/m} (B)\) for any Borelian \(B\) in \((0,1]\). Now, let \(\{N_ n\}\) be a sequence of positive integer-valued random variables such that \(\lim_{n \to \infty} N_ n/n = N\) in probability with a positive random variable \(N\). Assuming that \(\{X_ n\}\) is a \(T\)- periodic sequence with an external index \(\theta > 0\) and satisfying a long range dependence condition the limit of \(P(S_{N_ n} [X,u_{N_ n} (x)] ((0,b]) \leq s, B)\) is calculated. This problem has been discussed for i.i.d. sequences by \textit{J. Galambos} [``The asymptotic theory of extreme order statistics'' (1978; Zbl 0381.62039)] and \textit{H. M. Barakat} and \textit{M. A. El-Shandidy} [Commentat. Math. Univ. Carol. 31, No. 2, 323-336 (1990; Zbl 0721.60028) and J. Stat. Plann. Inference 26, No. 3, 353-361 (1990; Zbl 0718.60020)].
    0 references
    point process of exceedances
    0 references
    sample of random size
    0 references
    periodic sequence of random variables
    0 references
    long range dependence condition
    0 references

    Identifiers