On maximum likelihood estimators of shape and scale parameters and their application in constructing confidence contours.

From MaRDI portal
Publication:3313110





In this report the maximum likelihood estimators of shape and scale parameters, \(\alpha\) and \(\beta\) respectively, for the survival distribution \(\bar F\) defined by \(\bar F(x)=R[(x/\beta)^{\alpha}]\) for \(x>0\) with R a prescribed reliability function, are shown to have the property that \(U={\hat \alpha}/\alpha\), \(V=({\hat \beta}/\beta)^{\alpha}\) have a joint distribution independent of \(\alpha\) and \(\beta\) under quite general assumptions on the censoring type. In particular, the statistic \(T=U \ell n V\) has a distribution depending upon R and the nature of censoring but not upon \(\alpha\) and \(\beta\). This result is applied to the problem of constructing upper and lower confidence contours for \(\bar F\) using the estimated function \(R[(x/{\hat \beta})^{{\hat \alpha}}]\). This is accomplished by utilizing pseudo- random number generators in a computer program which enables the statistician to compute the desired percentiles of the requisite distributions. This technique is illustrated in a fatigue life application when the censoring distribution has parameters with a known relationship to the unknown parameters of life. (From author's abstract.)











This page was built for publication: On maximum likelihood estimators of shape and scale parameters and their application in constructing confidence contours.

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q3313110)