×

zbMATH — the first resource for mathematics

On the choice of a prior for Bayesian D-optimal designs for the logistic regression model with a single predictor. (English) Zbl 1333.62185
Summary: The Bayesian design approach accounts for uncertainty of the parameter values on which optimal design depends, but Bayesian designs themselves depend on the choice of a prior distribution for the parameter values. This article investigates Bayesian D-optimal designs for two-parameter logistic models, using numerical search. We show three things: (1) a prior with large variance leads to a design that remains highly efficient under other priors, (2) uniform and normal priors lead to equally efficient designs, and (3) designs with four or five equidistant equally weighted design points are highly efficient relative to the Bayesian D-optimal designs.

MSC:
62K05 Optimal statistical designs
62K25 Robust parameter designs
PDF BibTeX XML Cite
Full Text: DOI
References:
[1] DOI: 10.1080/01621459.1983.10477936 · doi:10.1080/01621459.1983.10477936
[2] Atkinson A. C., Optimum Experimental Designs, with SAS (2007) · Zbl 1183.62129
[3] DOI: 10.1016/S0169-7161(96)13016-9 · doi:10.1016/S0169-7161(96)13016-9
[4] DOI: 10.1007/978-1-4757-4286-2 · doi:10.1007/978-1-4757-4286-2
[5] DOI: 10.1002/9780470746912 · Zbl 1182.62154 · doi:10.1002/9780470746912
[6] DOI: 10.1093/biomet/asr001 · Zbl 1215.62072 · doi:10.1093/biomet/asr001
[7] Braess D., Institute of Mathematical Statistics 35 pp 772– (2007)
[8] DOI: 10.1080/03610928908829973 · Zbl 0696.62335 · doi:10.1080/03610928908829973
[9] DOI: 10.1016/0378-3758(89)90004-9 · Zbl 0666.62073 · doi:10.1016/0378-3758(89)90004-9
[10] DOI: 10.1214/ss/1177009939 · doi:10.1214/ss/1177009939
[11] DOI: 10.1214/aoms/1177728915 · Zbl 0053.10504 · doi:10.1214/aoms/1177728915
[12] DOI: 10.1214/aos/1032526965 · Zbl 0866.62046 · doi:10.1214/aos/1032526965
[13] DOI: 10.1111/1467-9868.00056 · doi:10.1111/1467-9868.00056
[14] Dette H., South Africa Statistical Journal 31 pp 271– (1997)
[15] DOI: 10.1016/j.jspi.2007.06.017 · Zbl 1139.62332 · doi:10.1016/j.jspi.2007.06.017
[16] DOI: 10.1198/004017006000000318 · doi:10.1198/004017006000000318
[17] DOI: 10.1111/1467-9868.00096 · doi:10.1111/1467-9868.00096
[18] Ford I., Journal of the Royal Statistical Society: Series B 54 pp 569– (1992)
[19] DOI: 10.1198/TECH.2009.0009 · doi:10.1198/TECH.2009.0009
[20] DOI: 10.1111/j.0006-341X.2004.00148.x · Zbl 1130.62316 · doi:10.1111/j.0006-341X.2004.00148.x
[21] DOI: 10.1214/aos/1009210553 · Zbl 1012.62081 · doi:10.1214/aos/1009210553
[22] DOI: 10.1214/088342306000000105 · Zbl 1246.62168 · doi:10.1214/088342306000000105
[23] DOI: 10.1111/j.0006-341X.2000.01263.x · Zbl 1060.62545 · doi:10.1111/j.0006-341X.2000.01263.x
[24] DOI: 10.1214/aoms/1177728069 · Zbl 0073.14103 · doi:10.1214/aoms/1177728069
[25] Matlab, Handbook Matlab, Version 7.10.0.499 (R2010a) (2010)
[26] DOI: 10.1016/0378-3758(94)00042-T · Zbl 0812.62081 · doi:10.1016/0378-3758(94)00042-T
[27] Müller C., Metrika 48 pp 1– (1998)
[28] DOI: 10.1111/j.0006-341X.2002.00735.x · Zbl 1210.62103 · doi:10.1111/j.0006-341X.2002.00735.x
[29] DOI: 10.1081/BIP-120019267 · Zbl 1180.62182 · doi:10.1081/BIP-120019267
[30] DOI: 10.1016/S0378-3758(98)00078-0 · Zbl 0924.62081 · doi:10.1016/S0378-3758(98)00078-0
[31] DOI: 10.1007/978-94-009-5912-5 · doi:10.1007/978-94-009-5912-5
[32] DOI: 10.1080/03610929808832173 · Zbl 0901.62095 · doi:10.1080/03610929808832173
[33] DOI: 10.1016/j.csda.2008.04.037 · Zbl 1452.62584 · doi:10.1016/j.csda.2008.04.037
[34] DOI: 10.1198/004017005000000571 · doi:10.1198/004017005000000571
[35] DOI: 10.1198/TECH.2011.09197 · doi:10.1198/TECH.2011.09197
[36] DOI: 10.1080/01621459.1985.10478213 · doi:10.1080/01621459.1985.10478213
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. It attempts to reflect the references listed in the original paper as accurately as possible without claiming the completeness or perfect precision of the matching.