Main Article Content

Abstract

In Industrial and Pharmaceutical experiments it is desired to have best predictions of the response on the basis of small amount of data.  Mixture experiment generally aims to predict the response(s) for all possible mixture blends. When we compute optimal design for mixture response surface we must focus on prediction capability of the design. The conventional optimal criteria, such as D-, A- and E-optimality are not suitable for determining the prediction capability of designs. As I-optimal design minimizes the average variance of prediction over the mixture region, so it clearly focuses on prediction capability of the design. Hence I-optimal criterion seems to be more appropriate in this conjecture.  In this paper we propose the construction of I-optimal mixture designs for a quadratic Scheffé’s and Darroch and Waller’s model in three and four components, using two orthogonal blocks.  I-efficiency of designs is compared with the I-efficiency of D-optimal designs for Scheffé’s and Darroch and Waller’s models.

Keywords

Mixture Experiments Latin Squares Orthogonal Blocks I-optimality Average Pridiction Variance D-optimality

Article Details

Author Biography

Syed Adil Hussain, University of Gujrat, Pakistan

Department of Statistics, Lecturer

How to Cite
Hasan, T., & Hussain, S. A. (2022). I-optimal Designs for three and four component mixture models in orthogonal blocks. Pakistan Journal of Statistics and Operation Research, 18(4), 943-954. https://doi.org/10.18187/pjsor.v18i4.3369

References

  1. Becker NG, (1968). Models for the response of a mixture. Journal of the Royal Statistical Society, Series B, 30, 349–358. DOI: https://doi.org/10.1111/j.2517-6161.1968.tb00735.x
  2. Czitrom V (1988.) Mixture experiments with process variables: D-optimal orthogonal experimental designs. Communication in Statistics: Theory and Methods, 17(1), 105-121. DOI: https://doi.org/10.1080/03610928808829613
  3. Czitrom V (1989) Experimental Design for Four Mixture Components with Process Variables. Communication in Statistics: Theory and Methods 18(12), 4561–4581. DOI: https://doi.org/10.1080/03610928908830175
  4. Chan LY, Sandhu MK (1999). Optimal orthogonal block designs for a quadratic mixture model for three components. Journal of Applied Statistics, 26(2),19–34. DOI: https://doi.org/10.1080/02664769922629
  5. Darroch JN, Waller J (1985). Additivity and interactions in three-component experiments with mixtures. Biometrika 72, 153-163. DOI: https://doi.org/10.1093/biomet/72.1.153
  6. Draper NR, Pukelsheim F (1997) Mixture Models based on Homogenous Polynomials. Journal of Statistical Planning and Inference 71, 303–311. DOI: https://doi.org/10.1016/S0378-3758(98)00012-3
  7. Draper NR, Prescott P, Lewis SM, Dean AM, John PWM, Tuck MG, (1993) Mixture designs for four components in orthogonal blocks. Technometrics 253(3): 268–276. DOI: https://doi.org/10.1080/00401706.1993.10485322
  8. Goos P, Jones B, Syafitri U (2016). I-optimal mixture designs. Journal of American Statistical Association, 111, 899-911. DOI: https://doi.org/10.1080/01621459.2015.1136632
  9. Ghosh S, Liu T (1999) Optimal Mixture Designs for Four Components in two Orthogonal Blocks. Journal of Statistical Planning and Inference, 78, 219–228. DOI: https://doi.org/10.1016/S0378-3758(98)00215-8
  10. Hasan T, Khan MF (2011) Nearly A- and E-optimal Orthogonally Blocked Designs for Scheffé’s Quadratic Mixture Model with Three Components. Proceedings of Pakistan Academy of Sciences, 48(2),101-106.
  11. Hasan T, Khan MF (2012) Nearly Optimal Orthogonally Blocked Designs for Second Degree K-model in Three Components. Pakistan Journal of Statistics, 28(2), 177-187.
  12. John PWM (1984). Experiments with Mixtures Involving Process Variables. Technical Report 8 Center for Statistical Sciences University of Texas Austin TX, 1-17.
  13. Kiefer JC, Wolfowitz J (1959). Optimum Designs in Regression Problems. The Annals of Mathematical Statistics, 30(2): 271-294. DOI: https://doi.org/10.1214/aoms/1177706252
  14. Lambrakis D (1968b). Experiments with p-component mixtures. Journal of the Royal Statistical Society, DOI: https://doi.org/10.1111/j.2517-6161.1968.tb01512.x
  15. Series B, 30(1), 137–144.
  16. Laake P (1975). On the optimal allocation of observation in experiments with mixtures. Scandinavian Journal of Statistics: Theory and Applications, 2, 153–157.
  17. Liu S, Neudecker H (1995) A V-optimal design for Scheffe’s polynomial model. Statistics and Probability Letters, 23, 253–258. DOI: https://doi.org/10.1016/0167-7152(94)00122-O
  18. Nigam AK (1970) Block designs for mixture experiments. The Annals of Mathematical Statistics, 41, 1861-1869. DOI: https://doi.org/10.1214/aoms/1177696687
  19. Prescott P (1998) Nearly Optimal Orthogonally Blocked Designs for a Quadratic Mixture Model with q Components. Communication in Statistics: Theory and Methods, 27(10), 259-2580. DOI: https://doi.org/10.1080/03610929808832243
  20. Singh P (2003) Optimal Orthogonal Designs in Two Blocks for Darroch and Waller’s Quadratic Mixture Model in three and four Components. Metron LXI , 419–430.
  21. Sinha BK, Mandal NK, Pal M, Das P (2014) Optimal Mixture Experiments. New Delhi Springer. DOI: https://doi.org/10.1007/978-81-322-1786-2
  22. Scheffé H (1958) Experiments with Mixtures. Journal of the Royal Statistical Society, Series B, 20, 344-360. DOI: https://doi.org/10.1111/j.2517-6161.1958.tb00299.x