This paper searches $ A $-optimal designs for mixture polynomial models when the errors are heteroscedastic. Sufficient conditions are given so that $ A $-optimal designs for the complex mixture polynomial models can be constructed from the direct sum of $ A $-optimal designs for their sub-mixture models with different structures of heteroscedasticity. Several examples are presented to further illustrate and check optimal designs based on $ A $-optimality criterion.
Citation: Fei Yan, Junpeng Li, Haosheng Jiang, Chongqi Zhang. $ A $-Optimal designs for mixture polynomial models with heteroscedastic errors[J]. AIMS Mathematics, 2023, 8(11): 26745-26757. doi: 10.3934/math.20231369
This paper searches $ A $-optimal designs for mixture polynomial models when the errors are heteroscedastic. Sufficient conditions are given so that $ A $-optimal designs for the complex mixture polynomial models can be constructed from the direct sum of $ A $-optimal designs for their sub-mixture models with different structures of heteroscedasticity. Several examples are presented to further illustrate and check optimal designs based on $ A $-optimality criterion.
| [1] |
H. Scheffé, Experiments with mixtures, J. R. Stat. Soc. Ser. B, 20 (2018), 344–360. https://doi.org/10.1111/j.2517-6161.1959.tb00335.x doi: 10.1111/j.2517-6161.1959.tb00335.x
|
| [2] |
H. Scheffé, The simplex-centroid design for experiments with mixtures, J. R. Stat. Soc. Ser. B, 25 (1963), 235–251. https://doi.org/10.1111/j.2517-6161.1963.tb00506.x doi: 10.1111/j.2517-6161.1963.tb00506.x
|
| [3] |
L. Y. Chan, Optimal design for experiment with mixtures: A survey, Comm. Statist. Theory Methods, 29 (2000), 2281–2312. https://doi.org/10.1080/03610920008832607 doi: 10.1080/03610920008832607
|
| [4] | J. A. Cornell, Experiments with mixtures: Designs, models, and the analysis of mixture data, In: Wiley series in probability and statistics, New York: Wiley & Sons, 2002. http://doi.org/10.1002/9781118204221 |
| [5] | M. N. L. Huang, H. L. Hsu, C. J. Chou, T. Klein, Model-robust $D$- and $A$-optimal designs for mixture experiments, Statist. Sinica, 19 (2009), 1055–1075. |
| [6] |
P. Goos, B. Jones, U. Syafitri, $I$-optimal design of mixture experiments, J. Amer. Statist. Assoc., 111 (2016), 899–911. https://doi.org/10.1080/01621459.2015.1136632 doi: 10.1080/01621459.2015.1136632
|
| [7] |
H. H. Hao, X. Y. Zhu, X. F. Zhang, C. Q. Zhang, $R$-optimal design of the second-order Scheffé mixture model, Statist. Probab. Lett., 173 (2021), 109069. https://doi.org/10.1016/j.spl.2021.109069 doi: 10.1016/j.spl.2021.109069
|
| [8] |
L. Ling, G. H. Li, X. Y. Zhu, C. Q. Zhang, $R$-optimal designs for second-order Scheffé model with qualitative factors, AIMS Mathematics, 7 (2022), 4540–4551. https://doi.org/10.3934/math.2022253 doi: 10.3934/math.2022253
|
| [9] |
H. Dette, M. Trampisch, A general approach to $D$-optimal designs for weighted univariate polynomial regression models, J. Korean Stat. Soc., 39 (2010), 1–26. https://doi.org/10.1016/j.jkss.2009.11.002 doi: 10.1016/j.jkss.2009.11.002
|
| [10] |
D. P. Wiens, P. F. Li, $V$-optimal designs for heteroscedastic regression, J. Statist. Plann. Inference, 145 (2014), 125–138. https://doi.org/10.1016/j.jspi.2013.09.007 doi: 10.1016/j.jspi.2013.09.007
|
| [11] |
C. Rodríguez, I. Ortiz, I. Martínez, $A$-optimal designs for heteroscedastic multifactor regression models, Comm. Statist. Theory Methods, 45 (2016), 757–771. https://doi.org/10.1080/03610926.2013.835419 doi: 10.1080/03610926.2013.835419
|
| [12] |
L. He, R. X. Yue, $R$-optimal designs for multi-factor models with heteroscedastic errors, Metrika, 80 (2017), 717–732. http://dx.doi.org/10.1007/s00184-017-0624-1 doi: 10.1007/s00184-017-0624-1
|
| [13] |
L. He, D. J. He, Bayesian and maximin optimal designs for heteroscedastic multi-factor regression models, Statist. Papers, 2022. https://doi.org/10.1007/s00362-022-01368-y doi: 10.1007/s00362-022-01368-y
|
| [14] |
W. K. Wong, $G$-optimal designs for multifactor experiments with heteroscedastic errors, J. Statist. Plann. Inference, 40 (1994), 127–133. http://doi.org/10.1016/0378-3758(94)90146-5 doi: 10.1016/0378-3758(94)90146-5
|
| [15] |
G. Montepiedra, W. K. Wong, A new design criterion when heteroscedasticity is ignored, Ann. Inst. Statist. Math., 53 (2001), 418–426. http://doi.org/10.1023/A:1012435125788 doi: 10.1023/A:1012435125788
|
| [16] |
C. Rodríguez, I. Ortiz, $D$-optimum designs in multi-factor models with heteroscedastic errors, J. Statist. Plann. Inference, 128 (2005), 623–631. https://doi.org/10.1016/j.jspi.2003.12.013 doi: 10.1016/j.jspi.2003.12.013
|
| [17] |
U. Graßhoff, H. Großmann, H. Holling, R. Schwabe, Design optimality in multi-factor generalized linear models in the presence of an unrestricted quantitative factor, J. Stat. Plann. Inference, 137 (2007), 3882–3892. https://doi.org/10.1016/j.jspi.2007.04.007 doi: 10.1016/j.jspi.2007.04.007
|
| [18] |
F. Yan, C. Q. Zhang, H. Peng, Optimal designs for additive mixture model with heteroscedastic errors, Comm. Statist. Theory Method, 46 (2017), 6401–6411. https://doi.org/10.1080/03610926.2015.1129414 doi: 10.1080/03610926.2015.1129414
|
| [19] | V. V. Fedorov, Theory of optimal experiments, New York: Academic Press, 1972. |
| [20] |
X. F. Zhang, Z. B. Zhu, C. Q. Zhang, Multi-stage differential evolution algorithm for constrained $D$-optimal design, AIMS Mathematics, 6 (2021), 2956–2969. https://doi.org/10.3934/math.2021179 doi: 10.3934/math.2021179
|
Figures(2) / Tables(1)
Fei Yan, Junpeng Li, Haosheng Jiang, Chongqi Zhang. $ A $-Optimal designs for mixture polynomial models with heteroscedastic errors[J]. AIMS Mathematics, 2023, 8(11): 26745-26757. doi: 10.3934/math.20231369
DownLoad: