Designs for generalized linear models with several variables and model uncertainty
Designs for generalized linear models with several variables and model uncertainty
Standard factorial designs may sometimes be inadequate for experiments that aim to estimate a generalized linear model, for example, for describing a binary response in terms of several variables. A method is proposed for finding exact designs for such experiments which uses a criterion that allows for uncertainty in the link function, the linear predictor or the model parameters, together with a design search. Designs are assessed and compared by simulation of the distribution of efficiencies relative to locally optimal designs over a space of possible models. Exact designs are investigated for two applications and their advantages over factorial and central composite designs are demonstrated.
binary response, d-optimality, logistic regression, robust design, simulation
284-292
Woods, D. C.
ae21f7e2-29d9-4f55-98a2-639c5e44c79c
Lewis, S. M.
a69a3245-8c19-41c6-bf46-0b3b02d83cb8
Eccleston, J. A.
f2f29edd-66f2-47e2-9f59-abb30a7bc615
Russell, K. G.
70cb8386-5ea4-4ba5-9f5f-e7b760f92dfc
2006
Woods, D. C.
ae21f7e2-29d9-4f55-98a2-639c5e44c79c
Lewis, S. M.
a69a3245-8c19-41c6-bf46-0b3b02d83cb8
Eccleston, J. A.
f2f29edd-66f2-47e2-9f59-abb30a7bc615
Russell, K. G.
70cb8386-5ea4-4ba5-9f5f-e7b760f92dfc
Woods, D. C., Lewis, S. M., Eccleston, J. A. and Russell, K. G.
(2006)
Designs for generalized linear models with several variables and model uncertainty.
Technometrics, 48 (2), .
(doi:10.1198/004017005000000571).
Abstract
Standard factorial designs may sometimes be inadequate for experiments that aim to estimate a generalized linear model, for example, for describing a binary response in terms of several variables. A method is proposed for finding exact designs for such experiments which uses a criterion that allows for uncertainty in the link function, the linear predictor or the model parameters, together with a design search. Designs are assessed and compared by simulation of the distribution of efficiencies relative to locally optimal designs over a space of possible models. Exact designs are investigated for two applications and their advantages over factorial and central composite designs are demonstrated.
Text
Designs_for_generalized_linear_models.pdf
- Version of Record
Restricted to Registered users only
Request a copy
Text
glm_techreport.pdf
- Accepted Manuscript
Text
19342-01.pdf
- Other
More information
e-pub ahead of print date: 8 February 2006
Published date: 2006
Keywords:
binary response, d-optimality, logistic regression, robust design, simulation
Organisations:
Statistics
Identifiers
Local EPrints ID: 15828
URI: http://eprints.soton.ac.uk/id/eprint/15828
ISSN: 0040-1706
PURE UUID: f566d8f0-b39f-4c02-a8e2-82e56bd1610b
Catalogue record
Date deposited: 14 Jun 2005
Last modified: 16 Mar 2024 03:14
Export record
Altmetrics
Contributors
Author:
S. M. Lewis
Author:
J. A. Eccleston
Author:
K. G. Russell
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
