Title :
Percentiles of pivotal ratios for the MLE of the parameters of a Weibull regression model
Author :
Bugaighis, Mohamed M.
Author_Institution :
Dept. of Math., Moravian Coll., Bethlehem, PA, USA
fDate :
3/1/1993 12:00:00 AM
Abstract :
The maximum likelihood procedure provides efficient estimators for regression parameters of the power-rule model associated with the two-parameter Weibull distribution. Percentiles of three pivotal ratios related to the power-rule model for a two-parameter Weibull distribution are examined. It is shown that the distributions of these pivotal ratios are closely related to appropriate t and χ2 distributions. The regression equations given provide approximations to the percentiles of the pivotal ratios in terms of t and χ 2 distributions. Such approximations are appropriate in several practical situations where some accuracy can be sacrificed for expediency and tractability
Keywords :
maximum likelihood estimation; parameter estimation; reliability theory; statistical analysis; MLE; Weibull regression model; maximum likelihood procedure; percentiles; pivotal ratios; power-rule model; reliability; two-parameter Weibull distribution; Educational institutions; Equations; Life testing; Maximum likelihood estimation; Power engineering and energy; Reliability engineering; Reliability theory; State estimation; Stress; Weibull distribution;
Journal_Title :
Reliability, IEEE Transactions on