POINT AND INTERVAL ESTIMATES OF THE SCALE PARAMETER OF WEIBULL POPULATIONS USING POINT AND INTERVAL ESTIMATES FOR EXPONENTIAL POPULATIONS.
AIR FORCE INST OF TECH WRIGHT-PATTERSON AFB OH SCHOOL OF ENGINEERING
Pagination or Media Count:
The linear unbiased estimator of the parameter sigma for exponential populations was given as sigma C sub m X sub m by Harter using a suitably chosen order statistic X sub m. Interval estimates for sigma were also given by Harter. It is shown here that for Weibull populations with a fixed shape parameter K a linear consistent estimator for the scale parameter theta is given by theta C sub m to the 1Kth powerY sub m using the Weibull order statistic Y sub m and the coefficient C sub m from Harters estimate for exponential populations. The estimator obtained here is compared with an unbiased estimator for the Weibull population obtained by Quayle using a single order statistic. The interval estimates obtained by Quayle are found to agree to five significant digits with those found here. The estimator was computed from n1 to n100 using the value of m chosen to be most efficient for exponential populations and for K0.50.54.01.06.0. Author
- Statistics and Probability
- Manufacturing and Industrial Engineering and Control of Production Systems