• • 上一篇    

一类二阶模型下中心复合设计的预测方差

严静1, 陈雪平1, 张敏珏1, 王晓迪2   

  1. 1. 江苏理工学院数理学院, 常州 213001;
    2. 中央财经大学统计与数学学院, 北京 102206
  • 收稿日期:2022-03-17 修回日期:2022-06-02 发布日期:2022-12-13
  • 通讯作者: 陈雪平, Email: chenxueping@jsut.edu.cn
  • 基金资助:
    国家自然科学基金项目(11971204), 江苏省自然科学基金(BK20200108), 江苏理工学院中吴青年创新人才支持计划.

严静, 陈雪平, 张敏珏, 王晓迪. 一类二阶模型下中心复合设计的预测方差[J]. 系统科学与数学, 2022, 42(11): 3107-3118.

Yan Jing, Chen Xueping, Zhang Minjue, Wang Xiaodi. Prediction Variance Analysis of a Class of Central Composite Designs Under the Second-Order Model[J]. Journal of Systems Science and Mathematical Sciences, 2022, 42(11): 3107-3118.

Prediction Variance Analysis of a Class of Central Composite Designs Under the Second-Order Model

Yan Jing1, Chen Xueping1, Zhang Minjue1, Wang Xiaodi2   

  1. 1. School of Mathematics and Physics, Jiangsu University of Technology, Changzhou 213001;
    2. School of Statistics and Mathematics, Central University of Finance and Economics, Beijing 102206
  • Received:2022-03-17 Revised:2022-06-02 Published:2022-12-13
预测能力是评价设计优劣、模型选择的一 个重要准则,通过预测方差函数可以准确获得在一定区域内的预测 精度.文章首先介绍了最新提出的基于正交表的中心复合设计以及基于极小极大准则的改进设计, 然后详细给出了两类设计在给定球面上的平均预测方差函数、最小预测方差函数和最大预测方差函数的解析表 达式,其仅为球距和设计参数的函数, 这一结果可以为实际工作者在预测分析时提供重要参考.
Prediction ability is an important criterion to evaluate the quality of design and model selection. The prediction accuracy in a certain area can be accurately obtained through the prediction variance function. This paper first introduces the recently proposed central composite design based on orthogonal-array and the modified designs based on Minimax criterion, and then gives the analytical expressions of the average predictive variance function, minimum predictive variance function and maximum predictive variance function of these two kinds of designs for a given sphere. The results show that they are only the functions of sphere distance and the design parameters. It can provide an important reference for practical workers in prediction.

MR(2010)主题分类: 

()
[1] Box G E, Wilson K B. On the experimental attainment of optimum conditions. Journal of the Royal Statistical Society: Series B $($Methodological$)$, 1951, 13(1): 1-38.
[2] Box G E, Hunter J S. Multi-factor experimental designs for exploring response surfaces. The Annals of Mathematical Statistics, 1957, 28(1): 195-241.
[3] Box M J, Draper N R. Factorial designs, the |X'X| criterion, and some related matters. Technometrics, 1971, 13(4): 731-742.
[4] Box G E P, Draper N R. Empirical Model Building and Response Surfaces. New York: John Wiley, 1987.
[5] Box G E P, Behnken D W. Some new three level designs for the study of quantitative variables. Technometrics, 1960, 2(2): 455-475.
[6] Draper N R, Lin D K J. Small response-surface designs. Technometrics, 1990, 32(2): 187-194.
[7] Morris M D. A class of three-level experimental designs for response surface modeling. Technometrics, 2000, 42(2): 111-121.
[8] Xu H, Jaynes J, Ding X. Combining two-level and three-level orthogonal arrays for factor screening and response surface exploration. Statistica Sinica, 2014, 24(1): 269-289.
[9] Giovannitti-Jensen A, Myers R H. Graphical assessment of the prediction capability of response surface designs. Technometrics, 1989, 31(2): 159-171.
[10] Borkowski J J. Spherical prediction-variance properties of central composite and Box-Behnken designs. Technometrics, 1995, 37(4): 399-410.
[11] Liang L, Anderson-Cook C, Robinson T, et al. Three-dimensional variance dispersion graphs for split-plot designs. Journal of Computational and Graphical Statistics, 2006, 15(4): 757-778.
[12] Park Y, Richardson D, Montgomery D, et al. Prediction variance properties of second-order designs for cuboidal regions. Journal of Quality Technology, 2005, 37(4): 253-266.
[13] Vining G G, Myers R H. A graphical approach for evaluating response surface designs in terms of the mean squared error of prediction. Technometrics, 1991, 33(3): 315-326.
[14] Trinca L A, Gilmour S G. Variance dispersion graphs for comparing blocked response surface designs. Journal of Quality Technology, 1998, 30(4): 314-327.
[15] Wesley W R, Simpson J R, Parker P A, et al. Maximum and minimum prediction variance for spherical and cuboidal split plot designs. Communications in Statistics — Theory and Methods, 2009, 38(13): 2251-2266.
[16] Huang H. Multi-drug combination designs with experiments in silico. Statistics and Its Interface, 2021, 14(4): 373-388.
[17] Frey D D, Engelhardt F, Greitzer E M. A role for “one factor at a time” experimentation in parameter designs. Research in Engineering Design, 2003, 14: 65-74.
[18] Bergman R S, Cox C W, DePriest D J, et al. Effect of process variations on incandescent lamp performance. Journal of the Illuminating Engineering Society, 2013, 19(2): 132-141.
[19] Shearer G, Tzoganakis C. Free radical hydrosilylation of polypropylene. Journal of Applied Polymer Science, 1998, 65(3): 439-447.
[20] Admassu W, Tom B. Feasibility of using natural fishbone apatite as a substitute for hydroxya-patite in remediating aqueous heavy metals. Journal of Hazardous Materials, 1999, 69(2): 187-196.
[21] 郭瑞, 张建方. 单因素轮换法的一个调查研究. 数理统计与管理, 2008, 27(3): 432-438. (Guo R, Zhang J F. A research on one-factor-at-a-time expeirmentation. Application of Statistics and Mangement, 2008, 27(3): 432-438.
[22] Zhou Y D, Xu H. Composite designs based on orthogonal arrays and definitive screening designs. Journal of the American Statistical Association, 2016, 112(520): 1675-1683.
[23] Chen X P, Guo B, Liu M Q, et al. Robustness of orthogonal-array based composite designs to missing data. Journal of Statistical Planning and Inference, 2018, 194: 15-24.
[24] Zhang X R, Liu M Q, Zhou Y D. Orthogonal uniform composite designs. Journal of Statistical Planning and Inference, 2020, 206: 100-110.
[25] Fang K T, Zhang Y T. Generalized Multivariate Analysis. Beijing and Berlin: Science Press and Springer-Verlag, 1990.
[26] Draper N R, Guttman I. An index of rotatability. Technometrics, 1988, 30(1): 105-111.
[27] 陈雪平,林金官,黄性芳,等.区组大小不等的主效应设计. 中国科学:数学, 2017, 47(6): 765-778. (Chen X P, Lin J G, Huang X F, et al. Main-effect plans with blocks of natural size. SCIENTIA SINICA Mathematica, 2017, 47(6): 765-778.)
[1] 马海南,陈雪平.  部分纯净主效应设计[J]. 系统科学与数学, 2020, 40(2): 243-251.
[2] 熊子康,刘力惟,宁建辉,覃红. 混料切片试验设计[J]. 系统科学与数学, 2020, 40(2): 262-274.
[3] 李博文,肖意可,贾锴,陈璇,段晓君,李修和. 约束离散空间的试验设计方案研究[J]. 系统科学与数学, 2020, 40(2): 275-280.
[4] 陈雪平,张应山. 强度2混合水平正交表交互作用的\混杂度量[J]. 系统科学与数学, 2012, 32(7): 901-907.
[5] 庞善起;张应山. 标准混合差集矩阵的构造[J]. 系统科学与数学, 2010, 30(3): 410-416.
阅读次数
全文


摘要