Bootstrapping weighted empirical processes that do not converge weakly

  • Soumendra Nath Lahiri

Research output: Contribution to journalArticlepeer-review

1 Scopus citations

Abstract

We show that the bootstrap method provides valid approximations to the sampling distribution of a weighted empirical process on D[0,1] even in the cases where it fails to converge weakly. Furthermore, the result is applied to construct valid bootstrap confidence sets in such pathological cases.

Original languageEnglish
Pages (from-to)295-302
Number of pages8
JournalStatistics and Probability Letters
Volume37
Issue number3
DOIs
StatePublished - Mar 16 1998

Keywords

  • Bootstrap
  • Confidence sets
  • Weak convergence
  • Weighted empirical process

Fingerprint

Dive into the research topics of 'Bootstrapping weighted empirical processes that do not converge weakly'. Together they form a unique fingerprint.

Cite this