D-optimal minimax regression designs on discrete design space

One classical design criterion is to minimize the determinant of the covariance matrix of the regression estimates, and the designs are called D-optimal designs. To reflect the nature that the proposed models are only approximately true, we propose a robust design criterion to study response surface designs. Both the variance and bias are considered in the criterion. In particular, D-optimal minimax designs are investigated and constructed. Examples are given to compare D-optimal minimax designs with classical D-optimal designs.

[1]  Douglas P. Wiens,et al.  Minimax designs for approximately linear models with AR(1) errors , 1999 .

[2]  Jane J. Ye,et al.  Existence and symmetry of minimax regression designs , 2007 .

[3]  Arthur B. Yeh,et al.  A two-stage strategy for the construction of D-optimal experimental designs , 1998 .

[4]  Atanu Biswas,et al.  An efficient design for model discrimination and parameter estimation in linear models , 2002 .

[5]  Julie Zhou,et al.  Discrete minimax designs for regression models with autocorrelated MA errors , 2007 .

[6]  Holger Dette,et al.  Optimal discrimination designs for multifactor experiments , 1997 .

[7]  Charles R. Johnson,et al.  Matrix analysis , 1985, Statistical Inference for Engineers and Data Scientists.

[8]  D. Wiens.  Minimax Robust Designs and Weights for Approximately Specified Regression Models With Heteroscedastic Errors , 1998 .

[9]  L. Haines.  The application of the annealing algorithm to the construction of exact optimal designs for linear-regression models , 1987 .

[10]  Douglas P. Wiens,et al.  Minimax designs for approximately linear regression , 1992 .

[11]  W. Bischoff,et al.  Optimal designs which are efficient for lack of fit tests , 2006, math/0611372.

[12]  W. J. Studden,et al.  Efficient $D_s$-Optimal Designs for Multivariate Polynomial Regression on the $q$-Cube , 1988 .

[13]  Holger Dette,et al.  Robust designs for multivariate polynomial regression on the d-cube , 1994 .

[14]  Wolfgang Bischoff,et al.  Lack-of-fit-efficiently optimal designs to estimate the highest coefficient of a polynomial with large degree , 2006 .

[15]  F. Pukelsheim.  Optimal Design of Experiments , 1993 .

[16]  Holger Dette,et al.  Robust designs for polynomial regression by maximizing a minimum of D- and D1-efficiencies , 2001 .

[17]  G. Box,et al.  A Basis for the Selection of a Response Surface Design , 1959 .

[18]  Holger Dette,et al.  Robustness properties of minimally-supported Bayesian D-optimal designs for heteroscedastic models , 2001 .

[19]  James L. Rosenberger,et al.  Experimental Designs for Model Discrimination , 1993 .

[20]  S. Gupta,et al.  Statistical decision theory and related topics IV , 1988 .

[21]  Wilfried Seidel,et al.  An algorithm for calculating Γ-minimax decision rules under generalized moment conditions , 2001 .

[22]  William I. Notz,et al.  OPTIMAL DESIGNS FOR REGRESSION MODELS WITH POSSIBLE BIAS , 1989 .

[23]  H. Dette.  Bayesian D-Optimal and Model Robust Designs in Linear Regression Models , 1993 .

[24]  Jane J. Ye,et al.  for misspecified regression models , 2003 .

[25]  W. Notz,et al.  Robust designs for nearly linear regression , 1982 .

[26]  Wolfgang Bischoff,et al.  Efficient lack of fit designs that are optimal to estimate the highest coefficient of a polynomial , 2006 .

[27]  D. Wiens,et al.  Integer-Valued, Minimax Robust Designs for Estimation and Extrapolation in Heteroscedastic, Approximately Linear Models , 2000 .

[28]  Andrej Pázman,et al.  Applications of necessary and sufficient conditions for maximin efficient designs , 1998 .

[29]  Julie Zhou.  INTEGER-VALUED, MINIMAX ROBUST DESIGNS FOR APPROXIMATELY LINEAR MODELS WITH CORRELATED ERRORS , 2001 .