The Annals of Statistics

Robust designs for polynomial regression by maximizing a minimum of D- and D1-efficiencies

Holger Dette and Tobias Franke

Full-text: Open access

PDF File (183 KB)

Abstract

In the common polynomial regression of degree m we determine the design which maximizes the minimum of the $D$-efficiency in the model of degree $m$ and the $D_1$-efficiencies in the models of degree $m-j,\dots, m +k$ ($j, k\ge 0$ given). The resulting designs allow an efficient estimation of the parameters in the chosen regression and have reasonable efficiencies for checking the goodness-of-fit of the assumed model of degree $m$ by testing the highest coefficients in the polynomials of degree $m-j,\dots, m +k$ .

Our approach is based on a combination of the theory of canonical moments and general equivalence theory for minimax optimality criteria. The optimal designs can be explicitly characterized by evaluating certain associated orthogonal polynomials.

Article information

Source
Ann. Statist., Volume 29, Number 4 (2001), 1024-1049.

Dates
First available in Project Euclid: 14 February 2002

Permanent link to this document
https://projecteuclid.org/euclid.aos/1013699990

Digital Object Identifier
doi:10.1214/aos/1013699990

Mathematical Reviews number (MathSciNet)
MR1869237

Zentralblatt MATH identifier
1012.62080

Subjects
Primary: 62K05: Optimal designs 33C45: Orthogonal polynomials and functions of hypergeometric type (Jacobi, Laguerre, Hermite, Askey scheme, etc.) [See also 42C05 for general orthogonal polynomials and functions]

Keywords
Minimax optimal designs robust design D-optimality D_1-optimality t-test, associated orthogonal polynomials

Citation

Dette, Holger; Franke, Tobias. Robust designs for polynomial regression by maximizing a minimum of D - and D 1 -efficiencies. Ann. Statist. 29 (2001), no. 4, 1024--1049. doi:10.1214/aos/1013699990. https://projecteuclid.org/euclid.aos/1013699990


Export citation

References

  • Abramowitz, M. and Stegun, I. A. (1964). Handbook of Mathematical Functions. Dover, New York.
  • Box, G. E. P. and Draper, N. R. (1959). A basis for the selection of a response surface design. J. Amer. Statist. Assoc. 54 622-654.
    Mathematical Reviews (MathSciNet): MR21:7584
    Zentralblatt MATH: 0116.36804
    Digital Object Identifier: doi:10.2307/2282542
    JSTOR: links.jstor.org
  • Chihara, T. S. (1978). An Introduction to Orthogonal Polynomials. Gordon and Breach, New York.
    Mathematical Reviews (MathSciNet): MR58:1979
    Zentralblatt MATH: 0389.33008
  • Dette, H. (1990). A generalization of Dand D1-optimal design in polynomial regression. Ann. Statist. 18 1784-1804.
    Mathematical Reviews (MathSciNet): MR91k:62074
    Zentralblatt MATH: 0714.62068
    Digital Object Identifier: doi:10.1214/aos/1176347878
    Project Euclid: euclid.aos/1176347878
  • Dette, H. (1995). Optimal designs for identifying the degree of a polynomial regression. Ann. Statist. 23 1248-1266.
    Mathematical Reviews (MathSciNet): MR97d:62151
    Zentralblatt MATH: 0847.62064
    Digital Object Identifier: doi:10.1214/aos/1176324708
    Project Euclid: euclid.aos/1176324708
  • Dette, H. and Studden, W. J. (1997). The Theory of Canonical Moments with Applications in Statistics, Probability and Analysis. Wiley, New York.
    Mathematical Reviews (MathSciNet): MR98k:60020
    Zentralblatt MATH: 0886.62002
  • Franke, T. (2000). Du nd D1-optimale Versuchspl¨ane unter Nebenbedingungen und gewichtete Maximin-Versuchspl¨ane bei polynomialer Regression. Dissertation, RuhrUniv. Bochum (in German).
  • Grosjean, C. C. (1986). The weight functions, generating functions and miscellaneous properties of the sequences of orthogonal polynomials of the second kind associated with the Jacobi and the Gegenbauer polynomials. J. Comp. Appl. Math. 18 259-307.
    Mathematical Reviews (MathSciNet): MR88b:33020
    Zentralblatt MATH: 0627.33008
    Digital Object Identifier: doi:10.1016/0377-0427(86)90001-4
  • Guest, P. G. (1958). The spacing of observations in polynomial regression. Ann. Math. Statist. 29 294-299.
    Mathematical Reviews (MathSciNet): MR20:1392
    Zentralblatt MATH: 0087.15303
    Digital Object Identifier: doi:10.1214/aoms/1177706730
    Project Euclid: euclid.aoms/1177706730
  • Hoel, P. G. (1958). Efficiency problems in polynomial estimation. Ann. Math. Statist. 29 1134-1145.
    Mathematical Reviews (MathSciNet): MR20:7359
    Zentralblatt MATH: 0094.14501
    Digital Object Identifier: doi:10.1214/aoms/1177706446
    Project Euclid: euclid.aoms/1177706446
  • Huber, P. J. (1975). Robustness and designs. In A Survey of Statistical Designs (J. N. Srivastava, ed.) 287-301. North Holland, Amsterdam.
    Mathematical Reviews (MathSciNet): MR53:1842
    Zentralblatt MATH: 0318.62035
  • Kiefer, J. C. (1974). General equivalence theory for optimum designs (approximate theory). Ann. Statist. 2 849-879.
    Mathematical Reviews (MathSciNet): MR50:8856
    Zentralblatt MATH: 0291.62093
    Digital Object Identifier: doi:10.1214/aos/1176342810
    Project Euclid: euclid.aos/1176342810
  • Kiefer, J. C. and Wolfowitz, J. (1959). Optimum designs in regression problems. Ann. Math. Statist. 30 271-294.
    Mathematical Reviews (MathSciNet): MR21:3079
    Digital Object Identifier: doi:10.1214/aoms/1177706252
    Project Euclid: euclid.aoms/1177706252
  • Lasser, R. (1994). Orthogonal polynomials and hypergroups II: the symmetric case. Trans. Amer. Math. Soc. 341 749-770.
    Mathematical Reviews (MathSciNet): MR94d:33005
    Zentralblatt MATH: 0804.42013
    Digital Object Identifier: doi:10.2307/2154581
  • Lau, T. S. (1983). Theory of canonical moments and its applications in polynomial regression I, II. Technical Reports 83-23, 83-24, Purdue Univ.
  • Lau, T. S. (1988). D-optimal designs on the unit q-ball. J. Statist. Plann. Inference 19 299-315.
    Mathematical Reviews (MathSciNet): MR89k:62093
    Zentralblatt MATH: 0850.62603
    Digital Object Identifier: doi:10.1016/0378-3758(88)90038-9
  • L¨auter, E. (1974). Experimental design in a class of models. Math. Oper. Statist. 5 379-398.
    Mathematical Reviews (MathSciNet): MR55:13681
  • Pukelsheim, F. (1993). Optimal Design of Experiments. Wiley, New York.
    Mathematical Reviews (MathSciNet): MR94k:62124
  • Pukelsheim, F. and Studden, W. J. (1993). E-optimal designs for polynomial regression. Ann. Statist. 21 402-415.
    Mathematical Reviews (MathSciNet): MR94c:62109
    Zentralblatt MATH: 0787.62075
    Digital Object Identifier: doi:10.1214/aos/1176349033
    Project Euclid: euclid.aos/1176349033
  • Silvey, S. D. (1980). Optimal Design. Chapman and Hall, London.
    Mathematical Reviews (MathSciNet): MR82d:62123
  • Skibinsky, M. (1967). The range of the n + 1 th moment for distributions on 0 1. J. Appl. Probab. 4 543-552.
    Mathematical Reviews (MathSciNet): MR37:3624
    Digital Object Identifier: doi:10.2307/3212220
    JSTOR: links.jstor.org
  • Skibinsky, M. (1986). Principal representations and canonical moment sequences for distributions on an interval. J. Math. Anal. Appl. 120 95-120.
    Mathematical Reviews (MathSciNet): MR88c:60049
    Zentralblatt MATH: 0607.60013
    Digital Object Identifier: doi:10.1016/0022-247X(86)90207-6
  • Spruill, M. G. (1990). Good designs for testing the degree of a polynomial mean. Sankhy¯a Ser. B 52 67-74.
    Mathematical Reviews (MathSciNet): MR93e:62188
    Zentralblatt MATH: 0733.62075
  • Stigler, S. (1971). Optimal experimental design for polynomial regression. J. Amer. Statist. Assoc. 66 311-318.
    Zentralblatt MATH: 0217.51701
  • Studden, W. J. (1968). Optimal designs on Tchebycheff points. Ann. Math. Statist. 39 1435-1447.
    Mathematical Reviews (MathSciNet): MR37:7050
    Zentralblatt MATH: 0174.22404
    Project Euclid: euclid.aoms/1177698123
  • Studden, W. J. (1980). Ds-optimal designs for polynomial regression using continued fractions. Ann. Statist. 8 1132-1141. Studden, W. J. (1982a). Optimal designs for weighted polynomial regression using canonical moments. In Statistical Decision Theory and Related Topics III 2 (S. S. Gupta and J. O. Berger, eds.) 335-350. Academic Press, New York. Studden, W. J. (1982b). Some robust-type D-optimal designs in polynomial regression. J. Amer. Statist. Assoc. 77 916-921.
    Mathematical Reviews (MathSciNet): MR585711
    Zentralblatt MATH: 0447.62070
    Digital Object Identifier: doi:10.1214/aos/1176345150
    Project Euclid: euclid.aos/1176345150
  • Studden, W. J. (1989). Note on some p-optimal designs for polynomial regression. Ann. Statist. 17 618-623.
    Mathematical Reviews (MathSciNet): MR90g:62185
    Zentralblatt MATH: 0675.62047
    Digital Object Identifier: doi:10.1214/aos/1176347129
    Project Euclid: euclid.aos/1176347129
  • Szeg ¨o, G. (1975). Orthogonal Polynomials. Amer. Math. Soc. Colloq. Publ., Providence, RI.
    Mathematical Reviews (MathSciNet): MR51:8724
  • Wiens, D. P. (1992). Minimax designs for approximately linear regression. J. Statist. Plann. Inference 31 353-371.
    Mathematical Reviews (MathSciNet): MR93h:62131
    Digital Object Identifier: doi:10.1016/0378-3758(92)90142-F
  • Wong, W. K. (1994). Comparing robust properties of A-, D-, Eand G-optimal designs. Comput. Statist. Data Anal. 18 441-448.
    Mathematical Reviews (MathSciNet): MR1304793

The Institute of Mathematical Statistics

The Annals of Statistics

New content alerts

Email RSS ToC RSS Article
  • You have access to this content.
  • You have partial access to this content.
  • You do not have access to this content.
Turn Off MathJax

What is MathJax?

More like this

+ See more