The convergence rate of fixed width sequential confidence intervals for a probability density function (Q3759758)

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 4009600
Language Label Description Also known as
default for all languages
No label defined
    English
    The convergence rate of fixed width sequential confidence intervals for a probability density function
    scientific article; zbMATH DE number 4009600

      Statements

      The convergence rate of fixed width sequential confidence intervals for a probability density function (English)
      0 references
      0 references
      1987
      0 references
      rate of convergence
      0 references
      recursive estimator
      0 references
      kernel functions
      0 references
      stopping rules
      0 references
      sequential confidence interval of fixed width
      0 references
      point estimators of kernel type
      0 references
      Let \(X_ 1,X_ 2,..\). be i.i.d. and real valued, with unknown density f. For given real number x, f(x) is to be estimated by a sequential confidence interval of fixed width 2d and confidence coefficient 1- \(\alpha\). Consider the interval \(I_{n,d}(x)=[f_ n(x)-d,f_ n(x)+d]\), in which \(f_ n\) is a sequence of point estimators of kernel type. The following type of stopping time is \(chosen:\) N\({}_ d\) is the least integer \(n\geq 1\) such that \(cn^{1-r} d^ 2\geq f_ n(x)+n^{-1}\), in which c and r depend on \(\alpha\) and the kernel. Then it is shown that \[ P\{f(x)\in I_{N_ d,d}\}=\alpha +O(d^{\eta})\quad as\quad d\to 0, \] in which \(\eta >0\) is an explicitly given function of r.
      0 references

      Identifiers