An exposition on the viability of the jackknife statistics in the estimation of the truncation point of continuous distributions

Date of Publication

1989

Document Type

Bachelor's Thesis

Degree Name

Bachelor of Science in Applied Mathematics

College

College of Science

Department/Unit

Mathematics and Statistics

Abstract/Summary

This paper establishes estimators of the truncation point of continuous distribution using the jackknifing methodology. These estimators are the jackknife estimator, the first, second, and third order generalized jackknife estimators, that is, J(8), G , 1(8), G,2(8), G3(8), respectively. Calculations reveal that these estimators are able to reduce bias. J(8) and G1(8) are examined in terms of the mean-squared error (MSE). Only these estimators are examined for this case because of the tediousness of the procedure. Calculations reveal that these estimators are able to reduce the mean-squared error. The estimators established using the jackknife procedure were then applied to the truncated Cauchy and Exponential distributions using Monte Carlo simulations. Results show that these estimators produced estimates which are closer in value to the point of truncation. However, though the bias has been reduced, there was a corresponding increase in variance. Nonetheless, for J(8) and G1(8), although variance has increased, the mean-squared error decreased.

Abstract Format

html

Language

English

Format

Print

Accession Number

TU05684

Shelf Location

Archives, The Learning Commons, 12F, Henry Sy Sr. Hall

Physical Description

71 leaves

Keywords

Statistics; Distribution (Probability theory)

This document is currently not available here.

Share

COinS