Advances and Applications in Statistics
Volume 43, Issue 2, Pages 107 - 118
(December 2014)
|
|
ESTIMATION OF A CHANGE-POINT IN THE WEIBULL REGRESSION MODEL
Oscar Palmeros-Rojas, José A. Villaseñor-Alva and Francisco S. Tajonar-Sanabria
|
Abstract: In some lifetime studies abrupt changes in the hazard function are observed due to overhauls, major operations or specific maintenance activities. In such situations it is of interest to detect the location in time where such changes occur. The Weibull regression hazard model with a single change-point is proposed, which is a generalization of the well-known exponential model. Statistical inference is focused on the problem concerning the change-point, taking into account covariates and random censorship. A likelihood-based approach is used for obtaining estimators for a change-point as well as for the hazard and regression parameters for the proposed model. Then the main goal is to prove the consistency of these estimators by relying on modern empirical process theory. |
Keywords and phrases: empirical processes, Glivenko-Cantelli and Donsker classes, hazard function, maximum likelihood estimators, consistency. |
|
Number of Downloads: 357 | Number of Views: 1318 |
|