Maximum Likelihood Estimation from Renewal Testing

Abstract

The report considers the maximum likelihood estimates of life-time distributions over an interval (O,T) from the following time truncated experiment. At time zero, the beginning of the testing n(1 <or= n < infinity) items are put on test. When an item fails it is replaced and at time T all testing is stopped. Assumptions about the form of the life-time distribution on (O,T) are required. Distributions considered are: A single parameter class which includes the Weibull family; A multiple parameter class with increasing failure rate on (O,T); and A nonparametric class which includes the increasing failure rate family. Useful and desirable properties of the maximum likelihood estimates are shown.

Open PDF

Document Details

Document Type
Technical Report
Publication Date
Apr 01, 1972
Accession Number
AD0757168

Entities

People

  • Larry H. Crow

Organizations

  • United States Army Materiel Systems Analysis Activity

Tags

Communities of Interest

  • Materials and Manufacturing Processes
  • Weapons Technologies

DTIC Thesaurus Topics

  • Asymptotic Normality
  • Computational Science
  • Consistency
  • Data Science
  • Distribution Functions
  • Equations
  • Estimators
  • Information Science
  • Intervals
  • Maximum Likelihood Estimation
  • Normality
  • Probability
  • Random Variables
  • Statistical Algorithms
  • Statistics
  • Systems Analysis
  • Theorems

Fields of Study

  • Mathematics

Readers

  • Statistical inference.