Advertisement

Optimal exact designs of experiments via Mixed Integer Nonlinear Programming

  • Belmiro P. M. DuarteEmail author
  • José F. O. Granjo
  • Weng Kee Wong
Article

Abstract

Optimal exact designs are problematic to find and study because there is no unified theory for determining them and studying their properties. Each has its own challenges and when a method exists to confirm the design optimality, it is invariably applicable to the particular problem only. We propose a systematic approach to construct optimal exact designs by incorporating the Cholesky decomposition of the Fisher Information Matrix in a Mixed Integer Nonlinear Programming formulation. As examples, we apply the methodology to find D- and A-optimal exact designs for linear and nonlinear models using global or local optimizers. Our examples include design problems with constraints on the locations or the number of replicates at the optimal design points.

Keywords

Model-based optimal designs Exact designs Constrained designs Mixed Integer Nonlinear Programming Global Optimization 

Mathematics Subject Classification

62K05 90C47 

Notes

Acknowledgements

The research of Wong is partially supported by a grant from the National Institute of General Medical Sciences of the National Institutes of Health under Award Number R01GM107639. The content is solely the responsibility of the authors and does not necessarily represent the official views of the National Institutes of Health. The authors acknowledge two anonymous reviewers that contributed undoubtedly to improve the quality of the paper.

Supplementary material

11222_2019_9867_MOESM1_ESM.pdf (108 kb)
Supplementary material 1 (pdf 107 KB)

References

  1. Atkinson, A., Donev, A.: The construction of exact \(D\)-optimum experimental designs with application to blocking response surface designs. Biometrika 76(3), 515–526 (1989)MathSciNetzbMATHCrossRefGoogle Scholar
  2. Atkinson, A., Donev, A., Tobias, R.: Optimum Experimental Designs, with SAS. Oxford University Press, Oxford (2007)zbMATHGoogle Scholar
  3. Berman, S.: An extension of the arc sine law. Ann. Math. Stat. 33(2), 681–684 (1962)MathSciNetzbMATHCrossRefGoogle Scholar
  4. Boer, E., Hendrix, E.: Global optimization problems in optimal design of experiments in regression models. J. Glob. Optim. 18, 385–398 (2000)MathSciNetzbMATHCrossRefGoogle Scholar
  5. Box, G., Hunter, W.: The experimental study of physical mechanisms. Technometrics 7(1), 23–42 (1965)CrossRefGoogle Scholar
  6. Chang, F.C., Imhof, L., Sun, Y.Y.: Exact \(D\)-optimal designs for first-order trigonometric regression models on a partial circle. Metrika Int. J. Theor. Appl. Stat. 76(6), 857–872 (2013)MathSciNetzbMATHGoogle Scholar
  7. Chang, F.C., Yeh, Y.R.: Exact \(A\)-optimal designs for quadratic regression. Stat. Sin. 8, 527–534 (1998)MathSciNetzbMATHGoogle Scholar
  8. Chernoff, H.: Locally optimal designs for estimating parameters. Ann. Math. Stat. 24, 586–602 (1953)MathSciNetzbMATHCrossRefGoogle Scholar
  9. Coale, A., McNeil, D.: The distribution by age of the frequency of first marriage in a female cohort. J. Am. Stat. Assoc. 67(340), 743–749 (1972)CrossRefGoogle Scholar
  10. Cook, R., Nachtsheim, C.: Comparison of algorithms for constructing \(D\)-optimal design. Technometrics 22(3), 315–324 (1980)zbMATHCrossRefGoogle Scholar
  11. Donev, A.: Crossover designs with correlated observations. J. Biopharm. Stat. 8(2), 249–262 (1998). PMID: 9598421zbMATHCrossRefGoogle Scholar
  12. Drud, A.: CONOPT: a GRG code for large sparse dynamic nonlinear optimization problems. Math. Program. 31, 153–191 (1985)MathSciNetzbMATHCrossRefGoogle Scholar
  13. Du Croz, J., Higham, N.: Stability of methods for matrix inversion. IMA J. Numer. Anal. 12, 1–19 (1992)MathSciNetzbMATHCrossRefGoogle Scholar
  14. Duarte, B., Wong, W.: Finding Bayesian optimal designs for nonlinear models: a semidefinite programming-based approach. Int. Stat. Rev. 83(2), 239–262 (2015)MathSciNetCrossRefGoogle Scholar
  15. Duarte, B., Wong, W., Atkinson, A.: A semi-infinite programming based algorithm for determining \(T\)-optimum designs for model discrimination. J. Multivar. Anal. 135, 11–24 (2015)MathSciNetzbMATHCrossRefGoogle Scholar
  16. Duarte, B., Wong, W., Dette, H.: Adaptive grid semidefinite programming for finding optimal designs. Stat. Comput. 28(2), 441–460 (2018)MathSciNetzbMATHCrossRefGoogle Scholar
  17. Duran, M., Grossmann, I.: An outer-approximation algorithm for a class of mixed-integer nonlinear programs. Math. Program. 36(3), 307–339 (1986)MathSciNetzbMATHCrossRefGoogle Scholar
  18. Esteban-Bravo, M., Leszkiewicz, A., Vidal-Sanz, J.: Exact optimal experimental designs with constraints. Stat. Comput. 27(3), 845–863 (2017)MathSciNetzbMATHCrossRefGoogle Scholar
  19. Fedorov, V.: Theory of Optimal Experiments. Academic Press, New York (1972)Google Scholar
  20. Fedorov, V., Leonov, S.: Optimal Design for Nonlinear Response Models. Chapman and Hall/CRC Press, Boca Raton (2014)zbMATHGoogle Scholar
  21. Fletcher, R., Leyffer, S.: Numerical experience with lower bounds for MIQP branch-and-bound. SIAM J. Optim. 8(2), 604–616 (1998)MathSciNetzbMATHCrossRefGoogle Scholar
  22. Floudas, C.: Mixed-integer nonlinear optimization. In: Pardalos, P., Resende, M. (eds.) Handbook of Applied Optimization, pp. 451–475. Oxford University Press, Oxford (2002)Google Scholar
  23. Gaffke, N.: On \(D\)-optimality of exact linear regression designs with minimum support. J. Stat. Plan. Inference 15, 189–204 (1987)MathSciNetzbMATHCrossRefGoogle Scholar
  24. GAMS Development Corporation: GAMS—a user’s guide, GAMS release 24.2.1. GAMS Development Corporation, Washington, DC, USA (2013a)Google Scholar
  25. GAMS Development Corporation: GAMS—the solver manuals, GAMS release 24.2.1. GAMS Development Corporation, Washington, DC, USA (2013b)Google Scholar
  26. Golub, G.H., van Loan, C.F.: Matrix Computations, 4th edn. JHU Press, Baltimore (2013)zbMATHGoogle Scholar
  27. Goos, P., Donev, A.: Blocking response surface designs. Comput. Stat. Data Anal. 51(2), 1075–1088 (2006)MathSciNetzbMATHCrossRefGoogle Scholar
  28. Goos, P., Vandebroek, M.: \(D\)-optimal response surface designs in the presence of random block effects. Comput. Stat. Data Anal. 37(4), 433–453 (2001)MathSciNetzbMATHCrossRefGoogle Scholar
  29. Goos, P., Vandebroek, M.: \(D\)-optimal split-plot designs with given numbers and sizes of whole plots. Technometrics 45(3), 235–245 (2003)MathSciNetCrossRefGoogle Scholar
  30. Gotwalt, C., Jones, B., Steinberg, D.: Fast computation of designs robust to parameter uncertainty for nonlinear settings. Technometrics 51(1), 88–95 (2009)MathSciNetCrossRefGoogle Scholar
  31. Gribik, P., Kortanek, K.: Equivalence theorems and cutting plane algorithms for a class of experimental design problems. SIAM J. Appl. Math. 32, 232–259 (1977)MathSciNetzbMATHCrossRefGoogle Scholar
  32. Harman, R., Filová, L.: Computing efficient exact designs of experiments using integer quadratic programming. Comput. Stat. Data Anal. 71, 1159–1167 (2014)MathSciNetzbMATHCrossRefGoogle Scholar
  33. Harman, R., Jurík, T.: Computing \(c\)-optimal experimental designs using the simplex method of linear programming. Comput. Stat. Data Anal. 53(2), 247–254 (2008)MathSciNetzbMATHCrossRefGoogle Scholar
  34. Hill, A.: The possible effects of the aggregation of the molecules of haemoglobin on its dissociation curves. J. Physiol. 40(Suppl.), 4–7 (1910)Google Scholar
  35. Imhof, L.: \(A\)-optimum exact designs for quadratic regression. J. Math. Anal. Appl. 228, 157–165 (1998)MathSciNetzbMATHCrossRefGoogle Scholar
  36. Imhof, L., Lopez-Fidalgo, J., Wong, W.K.: Efficiencies of rounded optimal approximate designs for small samples. Stat. Neerlandica 55(3), 301–318 (2001)MathSciNetzbMATHCrossRefGoogle Scholar
  37. Johnson, M., Nachtsheim, C.: Some guidelines for constructing exact \(D\)-optimal designs on convex design spaces. Technometrics 25, 271–277 (1983)MathSciNetzbMATHGoogle Scholar
  38. Kiefer, J.: General equivalence theory for optimum design (approximate theory). Ann. Stat. 2, 849–879 (1974)MathSciNetzbMATHCrossRefGoogle Scholar
  39. Laird, A.: Dynamics of tumor growth. Br. J. Cancer 18(3), 490–502 (1964)CrossRefGoogle Scholar
  40. Lastusilta, T., Bussieck, M., Westerlund, T.: Comparison of some high-performance MINLP solvers. Chem. Eng. Trans. 11, 125–130 (2007)Google Scholar
  41. Leszkiewicz, A.: Three essays on conjoint analysis: optimal design and estimation of endogenous consideration sets. Ph.D. thesis, Universidad Carlos III de Madrid (2014)Google Scholar
  42. López-Fidalgo, J., Tommasi, C., Trandafir, P.: Optimal designs for discriminating between some extensions of the Michaelis–Menten model. J. Stat. Plan. Inference 138(12), 3797–3804 (2008)MathSciNetzbMATHCrossRefGoogle Scholar
  43. Mandal, A., Wong, W.K., Yu, Y.: Algorithmic searches for optimal designs. In: Dean, A., Morris, M., Stufken, J., Bingham, D. (eds.) Handbook of Design and Analysis of Experiments, pp. 755–786. CRC Press, Boca Ratton (2015)Google Scholar
  44. Meyer, R., Nachtsheim, C.: The coordinate-exchange algorithm for constructing exact optimal experimental designs. Technometrics 37, 60–69 (1995)MathSciNetzbMATHCrossRefGoogle Scholar
  45. Mitchell, T.: An algorithm for the construction of \(D\)-optimal experimental designs. Technometrics 20, 203–210 (1974)MathSciNetzbMATHGoogle Scholar
  46. Mitchell, T., Miller Jr., F.: Use of Design Repair to Construct Designs for Special Linear Models. Technical Report, pp. 130–131, Oak Ridge National Laboratory (1970)Google Scholar
  47. Molchanov, I., Zuyev, S.: Steepest descent algorithm in a space of measures. Stat. Comput. 12, 115–123 (2002)MathSciNetCrossRefGoogle Scholar
  48. Nuñez Ares, J., Goos, P.: An integer linear programming approach to find trend-robust run orders of experimental designs. J. Qual. Technol. 51(1), 37–50 (2019).  https://doi.org/10.1080/00224065.2018.1545496 CrossRefGoogle Scholar
  49. Palhazi Cuervo, D., Goos, P., Sörensen, K.: Optimal design of large-scale screening experiments: a critical look at the coordinate-exchange algorithm. Stat. Comput. 26(1), 15–28 (2016)MathSciNetzbMATHCrossRefGoogle Scholar
  50. Pantelides, C.: The consistent initialization of differential-algebraic systems. SIAM J. Sci. Stat. Comput. 9(2), 213–231 (1988)MathSciNetzbMATHCrossRefGoogle Scholar
  51. Pukelsheim, F.: Optimal Design of Experiments. SIAM, Philadelphia (1993)zbMATHGoogle Scholar
  52. Rasch, D., Hendrix, E., Boer, E.: Replication-free optimal designs in regression analysis. Comput. Stat. 12, 19–52 (1997)zbMATHGoogle Scholar
  53. Sagnol, G.: Computing optimal designs of multiresponse experiments reduces to second-order cone programming. J. Stat. Plan. Inference 141(5), 1684–1708 (2011)MathSciNetzbMATHCrossRefGoogle Scholar
  54. Sagnol, G., Harman, R.: Computing exact \(D\)-optimal designs by mixed integer second order cone programming. Ann. Stat. 43(5), 2198–2224 (2015)MathSciNetzbMATHCrossRefGoogle Scholar
  55. Sahinidis, N.: BARON 14.3.1: global optimization of mixed-integer nonlinear programs, user’s manual. The Optimization Firm, LLC, Pittsburgh (2014)Google Scholar
  56. Sartono, B., Goos, P., Schoen, E.: Constructing general orthogonal fractional factorial split-plot designs. Technometrics 57(4), 488–502 (2015a)MathSciNetCrossRefGoogle Scholar
  57. Sartono, B., Schoen, E., Goos, P.: Blocking orthogonal designs with mixed integer linear programming. Technometrics 57(3), 428–439 (2015b)MathSciNetCrossRefGoogle Scholar
  58. SAS Institute Inc.: \(\text{JMP}^{\textregistered }\) 13 User Guide, 2nd edn. SAS Institute Inc, Cary (2017)Google Scholar
  59. Tawarlamani, M., Sahinidis, N.: Convexification and Global Optimization in Continuous and Mixed Integer Nonlinear Programming, 1st edn. Kluwer Academic Publishers, Dordrecht (2002)CrossRefGoogle Scholar
  60. Vandenberghe, L., Boyd, S.: Applications of semidefinite programming. Appl. Numer. Math. 29, 283–299 (1999)MathSciNetzbMATHCrossRefGoogle Scholar
  61. Vo-Thanh, N., Jans, R., Schoen, E., Goos, P.: Symmetry breaking in mixed integer linear programming formulations for blocking two-level orthogonal experimental designs. Comput. Oper. Res. 97, 96–110 (2018)MathSciNetzbMATHCrossRefGoogle Scholar
  62. Welch, W.: Branch-and-bound search for experimental designs based on \(D\)-optimality and other criteria. Technometrics 24(1), 41–48 (1982)MathSciNetzbMATHGoogle Scholar
  63. Westerlund, T., Pettersson, F.: An extended cutting plane method for solving convex MINLP problems. Comput. Chem. Eng. 19, 131–136 (1995)CrossRefGoogle Scholar
  64. Wynn, H.: The sequential generation of \(D\)-optimum experimental designs. Ann. Math. Stat. 41(5), 1655–1664 (1970)MathSciNetzbMATHCrossRefGoogle Scholar
  65. Yang, J., Mandal, A., Majumdar, D.: Optimal design for two-level factorial experiments with binary response. Stat. Sin. 22(2), 885–907 (2012)MathSciNetzbMATHCrossRefGoogle Scholar
  66. Yang, J., Mandal, A., Majumdar, D.: Optimal designs for \(2^k\) factorial experiments with binary response. Stat. Sin. 26, 381–411 (2016)zbMATHMathSciNetGoogle Scholar

Copyright information

© Springer Science+Business Media, LLC, part of Springer Nature 2019

Authors and Affiliations

  1. 1.Department of Chemical and Biological Engineering, Instituto Politécnico de CoimbraInstituto Superior de Engenharia de CoimbraCoimbraPortugal
  2. 2.Department of Chemical Engineering, CIEPQPFUniversity of CoimbraCoimbraPortugal
  3. 3.Department of Biostatistics, Fielding School of Public HealthUCLALos AngelesUSA

Personalised recommendations