Abstract

By a sequence of random vectors {ζk}, we can construct empirical distributions of the type Qn = (log n)-1 Σnk=1 δζk/k. Statements on the convergence of these or similar distributions with probability 1 to a limit distribution are called almost sure theorems. We propose several methods which permit us to easily deduce the almost sure limit theorems from the classical limit theorems, prove the invariance principle of the type "almost sure," and investigate the convergence of generalized moments. Unlike the majority of the preceding papers, where only convergence to the normal law is considered, our results may be applied in the case of limit distributions of the general type.

Original languageEnglish
Pages (from-to)254-272
Number of pages19
JournalTheory of Probability and its Applications
Volume44
Issue number2
Publication statusPublished - 1 Jan 1999

Scopus subject areas

  • Statistics and Probability
  • Statistics, Probability and Uncertainty

Fingerprint Dive into the research topics of 'On almost sure limit theorems'. Together they form a unique fingerprint.

Cite this