Comparison of Maximum Likelihood and Least Squares Cumulative Estimators for the Gompertz-Makeham Hazard Rate Function.
Abstract
This research investigates the use of the cumulative distribution function and least squares analysis for estimating the three parameters of the hazard function, b + k(e sup(-at)). The least squares cumulative (LSC) estimators of the parameters are obtained by solving a system of nonlinear equations using the gradient method of Fletcher-Powell. The small sample property of the estimator is analyzed in a two-way comparison. Through Monte-Carle simulation, sample data sets with five to twenty-five time-to-failure data points are generated by assuming parameter values. The failure times are used as inputs to the LSC program and the resulting estimates are compared to maximum likelihood estimates.
Document Details
- Document Type
- Technical Report
- Publication Date
- Apr 01, 1975
- Accession Number
- ADA009410
Entities
People
- Richard Earl Sackett
Organizations
- United States Army Materiel Command