ON THE MAXIMUM DEVIATION OF THE SAMPLE DENSITY.

Abstract

The maximum relative deviation of a sample density from the true one is shown to be relatively stable as the sample size tends to infinity provided both the sample and true densities are sufficiently regular. Conditions are given under which the sample density converges to the true density w.p. one uniformly on compact sets and uniformly on compact sets at a specified rate. The key to establishing this result is a theorem on the large deviations of independent, identically distributed bivariate random variables. (Author)

Document Details

Document Type
Technical Report
Publication Date
Apr 20, 1966
Accession Number
AD0636439

Entities

People

  • Michael Woodroofe

Organizations

  • Stanford University

Tags

DTIC Thesaurus Topics

  • Random Variables

Fields of Study

  • Mathematics

Readers

  • Mathematical Modeling and Probability Theory.
  • Regression Analysis.