Post-Configural Polysampling Pushback Performance.

Abstract

The pushback estimates are a preliminary data modification followed by the application of a robust statistic to the modified data. Application of configural polysampling techniques and use of the minimum attainable variance and maximum attainable polyefficiency derived from these techniques aid in fine-tuning the pushback estimates. The form of the pushback estimate shown by traditional Monte Carlo methods to perform well in comparison to a good biweight is modified and the performance is improved. (Author)

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Feb 01, 1982
Accession Number
ADA113778

Entities

People

  • Katherine Bell Krystinik

Organizations

  • Princeton University

Tags

DTIC Thesaurus Topics

  • Data Analysis
  • Data Mining
  • Data Science
  • Efficiency
  • Information Science
  • Monte Carlo Method
  • New Jersey
  • Observation
  • Order Statistics
  • Statistics
  • Universities

Readers

  • Regression Analysis.