Numerical construction of parameter maximin D-optimal designs for binary response models
Numerical construction of parameter maximin D-optimal designs for binary response models
For the binary response model, we determine optimal designs based on the D-optimal criterion which are robust with respect to misspecifications of the unknown parameters.
We propose a maximin approach and provide a numerical method to identify the best two point designs for the commonly applied link functions. This method is broadly applicable and can be extended to designs with a given number (\geq 2) of support points and further link functions. The results are illustrated for the logistic and probit model, for which several examples of maximin D-optimal designs are calculated explicitly by our method.
binary response model, robust optimal design, maximin D-optimality, bayesian D-optimality, prior distribution
127-161
Biedermann, Stefanie
fe3027d2-13c3-4d9a-bfef-bcc7c6415039
Dette, Holger
8c7b1c2e-3adc-45df-acfc-9e76509a228e
2005
Biedermann, Stefanie
fe3027d2-13c3-4d9a-bfef-bcc7c6415039
Dette, Holger
8c7b1c2e-3adc-45df-acfc-9e76509a228e
Biedermann, Stefanie and Dette, Holger
(2005)
Numerical construction of parameter maximin D-optimal designs for binary response models.
South African Statistical Journal, 39 (2), .
Abstract
For the binary response model, we determine optimal designs based on the D-optimal criterion which are robust with respect to misspecifications of the unknown parameters.
We propose a maximin approach and provide a numerical method to identify the best two point designs for the commonly applied link functions. This method is broadly applicable and can be extended to designs with a given number (\geq 2) of support points and further link functions. The results are illustrated for the logistic and probit model, for which several examples of maximin D-optimal designs are calculated explicitly by our method.
Text
numconst.pdf
- Other
More information
Published date: 2005
Keywords:
binary response model, robust optimal design, maximin D-optimality, bayesian D-optimality, prior distribution
Organisations:
Statistics
Identifiers
Local EPrints ID: 41830
URI: http://eprints.soton.ac.uk/id/eprint/41830
ISSN: 0038-271X
PURE UUID: edfc138e-8e48-4a70-9513-83014131a8d2
Catalogue record
Date deposited: 06 Oct 2006
Last modified: 16 Mar 2024 03:51
Export record
Contributors
Author:
Holger Dette
Download statistics
Downloads from ePrints over the past year. Other digital versions may also be available to download e.g. from the publisher's website.
View more statistics
