Abstract
Many practical experiments on mixtures (where the components sum to one) include additional lower or upper bounds on components, or on linear combinations of them. Usually theory cannot be used to obtain a good design, and algorithmic methods are necessary. Some of the available methods are discussed. Their performance is evaluated on some examples, and the form of the optimal design is investigated.
This is a preview of subscription content, log in via an institution to check access.
Similar content being viewed by others
References
Atkinson, A. C. (1996) The usefulness of optimum experimental designs. J. Roy. Statist. Soc., B, 57, 59–76 (discussion 95–111).
Atkinson, A. C. and Donev, A. N. (1992) Optimum experimental designs. Clarendon: Oxford.
Chan, L.-Y. (1995) A review on optimal design for mixture models. In L.-Y. Chan & M. C. Liu (eds), Five decades as a mathematician and educator – on the 80th birthday of Pro-fessor Wong, 45–88. World Scientific: Singapore.
Cornell, J. A. (1990) Experiments with mixtures. 2nd edn. Wiley: New York.
Crosier, R. B. (1984) Mixture experiments: geometry and pseu-docomponents. Technometrics, 26, 209–16.
Crosier, R. B. (1986) The geometry of constrained mixture ex-periments. Technometrics, 28, 95–102.
Gaffke, N. (1987) On D-optimality of exact linear regression de-signs with minimum support. J. Statist. Plann. Inference, 15, 189–204.
Hardin, R. H. and Sloane, N. J. A. (1993) A new approach to the construction of optimal designs. J. Statist. Plann. Inference, 37, 339–69.
Kiefer, J. (1961) Optimum designs in regression problems, II. Ann. Math. Statist., 32, 298–325.
Martin, R. J., Platts, L. M., Seddon, A. B. and Stillman, E. C. (1997) The development of a design and analysis strategy for a mixture experiment on glass durability. (Preprint).
Maths Works. (1992) Matlab version 3.5. The Maths Works Inc.: Natick, MA.
Mikaeili, F. (1993) D-optimum design for full cubic on the q-simplex. J. Statist. Plann. Inference, 35, 121–30.
Morris, M. D. and Mitchell, T. J. (1995) Exploratory designs for computational experiments. J. Statist. Plann. Inference, 43, 381–402.
Piepel, G. F., Anderson, C. M. and Redgate, P. E. (1993) Re-sponse surface designs for irregularly-shaped regions. Paper presented at The 37th Annual Fall Technical Conference, Rochester, New York. 23pp.
Snee, R. D. (1975) Experimental designs for quadratic models in constrained mixture spaces. Technometrics, 17, 149–59.
Snee, R. D. (1979) Experimental designs for mixture systems with multicomponent constraints. Commun. Statist.-Theor. Meth., 8, 303–26.
Snee, R. D. and Marquardt, D. W. (1974) Extreme vertices de-signs for linear mixture models. Technometrics, 16, 399–408.
Stat-Ease. (1993/7) Design-Expert version 4.0, 5.0: User's guide. Stat-Ease Inc.: Minneapolis.
Uranisi, H. (1964) Optimum design for the special cubic regres-sion on the q-simplex. Mathematical Reports, General Ed-ucation Dept., Kyushu University, No. 1, 7–12.
Rights and permissions
About this article
Cite this article
Martin, R.J., Bursnall, M.C. & Stillman, E.C. Efficient designs for constrained mixture experiments. Statistics and Computing 9, 229–237 (1999). https://doi.org/10.1023/A:1008978116760
Issue Date:
DOI: https://doi.org/10.1023/A:1008978116760