Geodesic
The Application of Minimax Fit in Linear Model and its Extension in Generalised Linear Models
Bulletin of the Malaysian Mathematical Society, Tome 24 (2001) no. 1 
Nor Aishah Hamzah; Daud Yahaya. The
                        Application of Minimax Fit in Linear Model and its Extension
                        in Generalised Linear Models. Bulletin of the Malaysian Mathematical Society, Tome 24 (2001) no. 1. http://geodesic.mathdoc.fr/item/BMMS_2001_24_1_a3/
@article{BMMS_2001_24_1_a3,
     author = {Nor Aishah Hamzah and Daud Yahaya},
     title = {The
                        {Application} of {Minimax} {Fit} in {Linear} {Model} and its {Extension
}                        in {Generalised} {Linear} {Models}},
     journal = {Bulletin of the Malaysian Mathematical Society},
     year = {2001},
     volume = {24},
     number = {1},
     url = {http://geodesic.mathdoc.fr/item/BMMS_2001_24_1_a3/}
}
TY  - JOUR
AU  - Nor Aishah Hamzah
AU  - Daud Yahaya
TI  - The
                        Application of Minimax Fit in Linear Model and its Extension
                        in Generalised Linear Models
JO  - Bulletin of the Malaysian Mathematical Society
PY  - 2001
VL  - 24
IS  - 1
UR  - http://geodesic.mathdoc.fr/item/BMMS_2001_24_1_a3/
ID  - BMMS_2001_24_1_a3
ER  - 
%0 Journal Article
%A Nor Aishah Hamzah
%A Daud Yahaya
%T The
                        Application of Minimax Fit in Linear Model and its Extension
                        in Generalised Linear Models
%J Bulletin of the Malaysian Mathematical Society
%D 2001
%V 24
%N 1
%U http://geodesic.mathdoc.fr/item/BMMS_2001_24_1_a3/
%F BMMS_2001_24_1_a3

Voir la notice de l'article provenant de la source BMMS

The minimax fitting procedure is employed in some robust parameter estimation in the normal linear model. In this paper, we shall reformulate the procedure by using the deviance residual measures instead of the usual residual measures for the robust fitting procedure. This extends the range of application from the normal distribution to a more general exponential family of distributions. Special cases of the exact solution for the single parameter case and its algorithm will be discussed.