An evaluation of change-point estimators for a sequence of normal observations with unknown parameters |
| |
| Authors: | Jorge Garza-Venegas Alvaro Cordero Franco María Temblador-Pérez Mario Beruvides |
| |
| Affiliation: | 1. Tecnológico de Monterrey - Engineering and Sciences School, Monterrey, Nuevo León, México;2. Universidad Autónoma de Nuevo León - Facultad de Ciencias Físico Matemáticas, San Nicolás de los Garza, Nuevo León, México;3. Tecnológico de Monterrey - Quality and Manufacturing Center, Monterrey, Nuevo León, México;4. Texas Tech University - Industrial Manufacturing, and Systems Engineering Department, Lubbock, Texas |
| |
| Abstract: | Performance of maximum likelihood estimators (MLE) of the change-point in normal series is evaluated considering three scenarios where process parameters are assumed to be unknown. Different shifts, sample sizes, and locations of a change-point were tested. A comparison is made with estimators based on cumulative sums and Bartlett's test. Performance analysis done with extensive simulations for normally distributed series showed that the MLEs perform better (or equal) in almost every scenario, with smaller bias and standard error. In addition, robustness of MLE to non-normality is also studied. |
| |
| Keywords: | Bartlett's test CUSUM MLE Robustness Unknown parameters |
|
|
