Sensitivity and Asymptotic Properties of Bayesian Reliability Estimates.

Abstract

In decision theoretic approaches to estimation problems, loss functions of the type L(theta, cap theta) = determinant theta - cap theta determinant to the alpa power, alpha > 0 are often employed; often alpha = 2. In many applications of reliability and life testing, such loss functions are inappropriate. Alternative loss functions which appear to be more suited to the intended application are proposed and Bayesian estimates of the exponential parameter are obtained for these. Asymptotic expansions of such estimators are given and compared with estimators given in the literature.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Apr 01, 1982
Accession Number
ADA116217

Entities

People

  • Andrew P. Soms
  • Bernard Harris

Organizations

  • University of Wisconsin–Madison

Tags

DTIC Thesaurus Topics

  • Asymptotic Series
  • Contracts
  • Data Science
  • Estimators
  • Information Science
  • Literature
  • Mathematics
  • Military Research
  • Order Statistics
  • Probability
  • Probability Density Functions
  • Reliability
  • Sensitivity
  • Statistical Analysis
  • Statistics
  • Theorems
  • United States

Fields of Study

  • Mathematics

Readers

  • Analytical Mechanics
  • Statistical inference.

Technology Areas

  • AI & ML
  • AI & ML - Bayesian Inference
  • AI & ML - Machine Learning Algorithms