Location and scale parameter estimation from randomly censored data |
| |
Authors: | RL Eubank VN LaRiccia |
| |
Institution: | 1. Department of Statistics , Southern Methodist University , Dallas, Texas, 75275;2. Department of Math. Sciences , University of Delaware , Newark, Delaware, 19711 |
| |
Abstract: | The problem of location and scale parameter estimation from randomly censored data is analyzed through use of a regression model for the Kaplan-Meier quantlle process. Continuous time regression techniques are employed to construct estimators that are both asymptotically normal and efficient. Estimators with a particularly simple form are obtained for the Koziol-Green model for random censorship. In the event of no censoring the regression model, and resulting estimators, reduce to those proposed by Parzen (1979 a, b). |
| |
Keywords: | quantile function estimation location and scale parameters random censoring reproducing kernel Hilbert space |
|
|