No label defined (Q3345498)

From MaRDI portal

!

This is the item page for this Wikibase entity, intended for internal use and editing purposes. Please use the normal view instead:

scientific article; zbMATH DE number 3879808
Language Label Description Also known as
default for all languages
No label defined
    English
    No label defined
    scientific article; zbMATH DE number 3879808

      Statements

      0 references
      1984
      0 references
      rate of convergence
      0 references
      random central limit theorem for U-statistics
      0 references
      In this paper, we consider the rate of convergence in the random central limit theorem for U-statistics, and get two results: Theorem 1: Suppose that the kernel \(\phi\) of U-statistics \(\{U_ n\}\) satisfies E \(\phi\) (X\({}_ 1,X_ 2)=0\), \(E| \phi (X_ 1,X_ 2)|^{2+\delta}<\infty\) \((0<\delta \leq 1)\), \(E^ 2\{E(\phi (X_ 1,X_ 2)| X_ 1)\}>0\). Suppose also that \(\{\nu_ n\}\) is a sequence of positive integer-valued r.v. satisfying \(P(| n^{- 1}\nu_ n-\tau | >n^{-a})\leq O(n^{-\delta /2})\). Here \(\tau\) is a positive constant, \(a=\delta\), if \(0<\delta \leq \sqrt{3}-1\), and \(=\delta +\delta (2+\delta)^{-1}\), if \(\sqrt{3}-1<\delta \leq 1\). Then \[ \sup_{x}| P(\sigma^{-1}_{\nu_ m}U_{\nu_ m}<x)-\Phi (x)| =O(n^{-\delta /2}),\quad where\quad \sigma^ 2_ n=EU^ 2_ n. \] Theorem 2: Define U-statistics as in theorem 1, with \(\delta\in (0,1/3)\), suppose that a sequence \(\{\nu_ n\}\) of positive- valued r.v. satisfies \(P(| n^{-1}\nu_ n-\lambda | >n^{- a})\leq O(n^{-\tau})\). Here \(a\in (3\delta /2\), 1/2], \(\tau =\delta (1- a)(2-3\delta)^{-1}\), suppose also that \(\lambda\) is a positive valued r.v. satisfying \(P(\lambda \leq n^{(3\delta -2a)/(2-3\delta)})+P(\lambda >\log n)=O(n^{-\tau})\), and \(\sup_{A\in {\mathcal B}_ n^{\infty}}\), \(B\in F| P(A| B)-P(A)| \leq \theta (n)\downarrow 0\), where \(\theta (n)=O(e^{-\beta n})\) for some \(\beta >0\), F and \({\mathcal B}_ n^{\infty}\) are the \(\sigma\)-fields generated by \(\lambda\) and \(\{X_ m,m\geq n\}\), respectively. Then \[ \sup_{x}| P(\sigma^{- 1}_{\nu_ n}U_{\nu_ n}<x)-\Phi (x)| =O(n^{-\min (2\beta \tau /\delta,\tau)}). \]
      0 references

      Identifiers