| Hu, F. and Kalb eisch, J.D. (2000). The estimating function bootstrap. The Canadian Journal of Statistics, 28, ???-???. |
....proposal to the linear regression model. The re calculation of the residual scale estimate and its impact on the overall computational cost of the method have not received much attention in the literature. Many proposals to re calculate robust estimates (for example Schucany and Wang, 1991; and Hu and Kalb eisch, 2000) ignore the fact that when the errors do not have a symmetric distribution the scale estimate is no longer asymptotically independent from the regression parameters estimates. Because of this, for the bootstrap procedure to be asymptotically correct we have to re calculate the scale estimate 4 ....
....Workstation. Note that due to the rst problem discussed above, even if we wait for the results of this run, they might not be reliable. The second problem mentioned above has received some attention in the literature. See for example Schucany and Wang (1991) Hu and Zidek (1995) Singh (1998) and Hu and Kalb eisch (2000). Note however that simultaneous consideration of these problems seems to be missing in the literature. In particular, the need to re calculate the scale with each bootstrap sample has not yet been studied in the robustness literature. Our basic idea is to bootstrap a re weighted representation of ....
Hu, F. and Kalb eisch, J.D. (2000). The estimating function bootstrap. The Canadian Journal of Statistics, 28, ???-???.
No context found.
Hu, F. and Kalbfleisch, J.D. (1999). The estimating function bootstrap. Submitted. Fisher Lecture of the 1999 Joint Statistical Meeting.
Online articles have much greater impact More about CiteSeer.IST Add search form to your site Submit documents Feedback
CiteSeer.IST - Copyright Penn State and NEC