Statistical inference for a class of life distributions |
| |
Authors: | Moshe Shaked |
| |
Institution: | The University of New Mexico , Albuquerque, New Mexico, 87131 |
| |
Abstract: | Life distributions with hazard rate functions of the form r(t) = Ag(t) + Bh(t) are considered. It is assumed that g(t) and h(t) are known and are independent of the unknown parameters A and B. The maximum likelihood estimators are studied for complete and censored samples. The estimation problem is reduced to a solution of one equation with one unknown parameter and it is observed that the solution is unique. The estimation procedure under the assumption of aging is also described. Some comments about the asymptotic variance-covariance matrix are given, and tests of hypo-theses are described for some cases. |
| |
Keywords: | hazard rate maximum likelihood estimators complete and censored samples IHR IHRA and NBU distributions convex cone linear hazard rate piecewise exponential |
|
|