×

Robust discrimination designs. (English) Zbl 1248.62132

Summary: We study the construction of experimental designs, the purpose of which is to aid in the discrimination between two possibly nonlinear regression models, each of which might be only approximately specified. A rough description of our approach is that we impose neighbourhood structures on each regression response and determine the members of these neighbourhoods which are least favourable in the sense of minimizing the Kullback-Leibler divergence. Designs are obtained which maximize this minimum divergence. Both static and sequential approaches are studied. We then consider sequential designs whose purpose is initially to discriminate, but which move their emphasis towards efficient estimation or prediction as one model becomes favoured over the other.

MSC:

62K25 Robust parameter designs
62L05 Sequential statistical design
62K05 Optimal statistical designs
62J02 General nonlinear regression
62B10 Statistical aspects of information-theoretic topics

Software:

nlmdl
PDFBibTeX XMLCite
Full Text: DOI

References:

[1] Atkinson, DT-optimum designs for model discrimination and parameter estimation, J. Statist. Planng Inf. 138 pp 56– (2008) · Zbl 1144.62054
[2] Atkinson, Planning experiments for discriminating between models (with discussion), J. R. Statist. Soc. B 36 pp 321– (1974) · Zbl 0291.62095
[3] Atkinson, The design of experiments for discriminating between two rival models, Biometrika 62 pp 57– (1975) · Zbl 0308.62071
[4] Atkinson, Optimal design: experiments for discriminating between several models, Biometrika 62 pp 289– (1975) · Zbl 0321.62085
[5] Biedermann, Optimal designs for testing the functional form of a regression via nonparametric estimation techniques, Statist. Probab. Lett. 52 pp 215– (2001) · Zbl 1012.62079
[6] Bischoff, Optimal designs which are efficient for lack of fit tests, Ann. Statist. 34 pp 2015– (2006) · Zbl 1246.62175
[7] Box, A basis for the selection of a response surface design, J. Am. Statist. Ass. 54 pp 622– (1959) · Zbl 0116.36804
[8] Cox, The Theory of the Design of Experiments (2000) · Zbl 1009.62061
[9] Dette, A comparison of sequential and non-sequential designs for discrimination between nested regression models, Biometrika 91 pp 165– (2004) · Zbl 1132.62343
[10] Dette, Optimal discrimination designs, Ann. Statist. (2008) · Zbl 1168.62066
[11] Fang, Integer-valued, minimax robust designs for estimation and extrapolation in heteroscedastic, approximately linear models, J. Am. Statist. Ass. 95 pp 807– (2000) · Zbl 0995.62068
[12] Fedorov, A Survey of Statistical Design and Linear Models (1975)
[13] Fedorov, Design of physical experiments, Fortsch. Phys. 16 pp 325– (1968)
[14] Ford, Recent advances in nonlinear experimental design, Technometrics 31 pp 49– (1989) · Zbl 0668.62048
[15] Gallant, Nonlinear Statistical Models (1987)
[16] Hill, A review of experimental design procedures for regression model discrimination, Technometrics 20 pp 15– (1978) · Zbl 0378.62067
[17] Hill, A joint design criterion for the dual problem of model discrimination and parameter estimation, Technometrics 10 pp 145– (1968)
[18] Hunter, Designs for discriminating between two rival models, Technometrics 7 pp 307– (1965)
[19] López-Fidalgo, An optimal experimental design criterion for discriminating between non-normal models, J. R. Statist. Soc. B 69 pp 231– (2007) · Zbl 1123.62056
[20] Pukelsheim, Experimental designs for model discrimination, J. Am. Statist. Ass. 88 pp 642– (1993) · Zbl 0773.62052
[21] Sinha, Robust sequential designs for nonlinear regression, Can. J. Statist. 30 pp 601– (2002) · Zbl 1018.62057
[22] Sinha, Festschrift for Constance van Eeden pp 233– (2003)
[23] Uciński, T-optimum designs for discrimination between two multiresponse dynamic models, J. R. Statist. Soc. B 67 pp 3– (2005) · Zbl 1060.62084
[24] Wiens, Designs for approximately linear regression: two optimality properties of uniform designs, Statist. Probab. Lett. 12 pp 217– (1991) · Zbl 0736.62066
[25] Wiens, Minimax designs for approximately linear regression, J. Statist. Planng Inf. 31 pp 353– (1992) · Zbl 0755.62058
[26] Wiens, Asymptotic properties of a Neyman-Pearson test for model discrimination, with an application to experimental design, J. Statist. Theory Pract. (2009) · Zbl 1211.62038 · doi:10.1080/15598608.2009.10411934
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. In some cases that data have been complemented/enhanced by data from zbMATH Open. This attempts to reflect the references listed in the original paper as accurately as possible without claiming completeness or a perfect matching.