實驗設計為資料分析的首要環節。在資料收集階段，根據實驗目的制定最佳化準則與設計資料收集方針，使得實驗設計能最有效地提供準確的統計推論結果。數學理論為實驗設計建構之一環，然而，理論建構之設計常具有限制，如僅適用於特定統計模型。實驗設計建構問題之根本為一最佳化問題，因此，數值方法為建構實驗設計不可或缺之工具。傳統數值建構設計方法為「替換」演算法。因其特性易收斂至局部最佳解，實務上的使用方式為隨機生成一群起始設計，各自迭代至收斂後，輸出其中最佳結果。本論文針對實驗設計建構之數值方法提出有別於傳統之想法，利用群集智慧演算法建構實驗設計，於一群起始設計中納入相互學習機制，更有效率地建構最佳實驗設計。文中，我們探討離散型實驗設計之建構問題，分別針對具模型辨識能力之兩水準因子設計與主作用與特定交互作用模型結構下的直交表，提出相對應之建構演算法。最後，我們介紹所開發之軟體供各領域之研究者操作使用。

The experimental design is critically important in the first stage of data analysis. To collect data, we defined the optimal criteria based on the experimental requirements and then generated the design accordingly. An optimal design offers accurate inferences at minimal cost. Theoretically proving a design's optimal structure is a typical approach to obtaining a proper design, but there are often restrictions. Such restriction can be on the run size of the optimal design and the structure of the statistic model. Essentially, we could treat design-construction problems as optimization problems and then use numerical algorithms to find the optimal design. The exchange-type algorithm is a commonly used tool for design construction. In practice, we ran this algorithm simultaneously many times with various initial designs and then output the best one of those results, as the exchange-type algorithm can easily become trapped in the local optimizers. For this reason, in this dissertation, we propose a novel way to efficiently construct optimal designs. This method involves a class of algorithms called swarm-intelligence algorithms, which allow the initial starters to learn from each other through a specialized learning mechanism. In this dissertation, we propose two swarm-intelligence algorithms related to discrete design, a two-level discrimination design for a set of linear models with main effects and two-factor interactions, and an orthogonal array for estimating the main effects and some prespecified two-factor interactions. In the end, we introduce our novel development of design-generating software for practical use.

[1] V. Agboto, W. W. Li, and C. Nachtsheim. Screening designs for model discrimination. Journal of Statistical Planning and Inference, 140(3):766–780, 2010.
[2] A. Aliev, V. Fedorov, S. Leonov, B. McHugh, and M. Magee. PkStaMp library for constructing optimal population designs for pk/pd studies. Communications in Statistics-Simulation and Computation, 41(6):717–729, 2012.
[3] A. C. Atkinson and D. R. Cox. Planning experiments for discriminating between models. Journal of the Royal Statistical Society. Series B (Methodological), 36(3):321–348, 1974.
[4] A. C. Atkinson and V. V. Fedorov. The design of experiments for discriminating between two rival models. Biometrika, 62(1):57–70, 1975.
[5] A. C. Atkinson and V. V. Fedorov. Optimal design: experiments for discriminating between several models. Biometrika, 62(2):289–303, 1975.
[6] S. Biedermann, H. Dette, and A. Pepelyshev. Maximin optimal designs for a compartmental model. In MODA 7 - Advances in Model-Oriented Design and Analysis, pages 41–49. Springer, 2004.
[7] B. Bogacka, M. Patan, P. J. Johnson, K. Youdim, and A. C. Atkinson. Optimum design of experiments for enzyme inhibition kinetic models. Journal of biopharmaceutical statistics, 21(3):555–572, 2011.
[8] G. E. P. Box and W. J. Hill. Discrimination among mechanistic models. Technometrics, 9(1):57–71, 1967.
[9] P.-Y. Chen, R.-B. Chen, H.-C. Tung, and W. K. Wong. Standardized maximim D-optimal designs for enzyme kinetic inhibition models. Chemometrics and Intelligent Laboratory Systems, 169:79–86, 2017.
[10] R.-B. Chen, S.-P. Chang, W. Wang, H.-C. Tung, and W. K. Wong. Minimax optimal designs via particle swarm optimization methods. Statistics and Computing, 25(5):975–988, 2015.
[11] R.-B. Chen, P.-Y. Chen, C.-L. Hsu, and W. K. Wong. Hybrid algorithms for generating optimal discrimination designs. Preprint, 2017.
[12] R.-B. Chen, D.-N. Hsieh, Y. Hung, and W. Wang. Optimizing latin hypercube designs by particle swarm. Statistics and Computing, 23(5):663–676, 2013.
[13] R.-B. Chen, Y.-W. Hsu, Y. Hung, and W. Wang. Discrete particle swarm optimization for constructing uniform design on irregular regions. Computational Statistics & Data Analysis, 72:282–297, 2014.
[14] C.-S. Cheng and B. Tang. A general theory of minimum aberration and its applications. The Annals of Statistics, 33(2):944–958, 2005.
[15] H. Chernoff. Locally optimal designs for estimating parameters. The Annals of Mathematical Statistics, pages 586–602, 1953.
[16] S. Consoli, J. A. Moreno-Pérez, K. Darby-Dowman, and N. Mladenović. Discrete particle swarm optimization for the minimum labelling steiner tree problem. Natural Computing, 9(1):29–46, 2010.
[17] R. D. Cook and C. J. Nachtrheim. A comparison of algorithms for constructing exact D-optimal designs. Technometrics, 22(3):315–324, 1980.
[18] R. D. Cook and C. J. Nachtsheim. An analysis of the k-exchange algorithm. Technical report, Tech. rep., University of Minnesota, Dept. of Operations and Management Science, 1982.
[19] A. Coppola and B. M. Stewart. lbfgs: Efficient L-BFGS and OWL-QN Optimization in R, 2014. R package version 1.2.1.
[20] L.-Y. Deng and B. Tang. Design selection and classification for hadamard matrices using generalized minimum aberration criteria. Technometrics, 44(2):173–184, 2002.
[21] H. Dette and S. Titoff. Optimal discrimination designs. The Annals of Statistics, 37(4):2056–2082, 2009.
[22] A. Dey and C.-Y. Suen. Optimal fractional factorial plans for main effects and specified two-factor interactions: a projective geometric approach. The Annals of Statistics, 30(5):1512–1523, 2002.
[23] K.-L. Du and M. N. S. Swamy. Search and Optimization by Metaheuristics. Springer, 2016.
[24] R. C. Eberhart and J. Kennedy. A new optimizer using particle swarm theory. In The 6th International Symposium on Micro Machine and Human Science, pages 39–43. IEEE, 1995.
[25] D. Eddelbuettel, D. Ardia, K. Mullen, B. Peterson, J. Ulrich, and R. Storn. RcppDE:Global Optimization by Differential Evolution in C++, 2016. R package version 0.1.5.
[26] D. Eddelbuettel, R. François, J. Allaire, J. Chambers, D. Bates, and K. Ushey. Rcpp:Seamless R and C++ integration. Journal of Statistical Software, 40(8):1–18, 2011.
[27] P. T. Eendebak and E. D. Schoen. Two-level designs to estimate all main effects and two-factor interactions. Technometrics, 59(1):69–79, 2017.
[28] V. V. Fedorov. Theory of Optimal Experiments. Elsevier, 1972.
[29] M. F. Franklin and R. A. Bailey. Selection of defining contrasts and confounded effects in two-level experiments. Journal of the Royal Statistical Society. Series C (Applied Statistics), 26(3):321–326, 1977.
[30] F. J. M. Garcia and J. A. M. Perez. Jumping frogs optimization: a new swarm method for discrete optimization. Documentos de Trabajo del DEIOC, 3, 2008.
[31] M. Gómez-González, A. López, and F. Jurado. Hybrid discrete PSO and OPF approach for optimization of biomass fueled micro-scale energy system. Energy Conversion and Management, 65:539–545, 2013.
[32] A. A. Greenfield. Selection of defining contrasts in two-level experiments. Journal of the Royal Statistical Society. Series C (Applied Statistics), 25(1):64–67, 1976.
[33] U. Groemping. Cran task view: Design of experiments (DoE) and analysis of experimental data. 2017. [Online; accessed 29-Nov-2017].
[34] M. Hamada and C. F. J. Wu. Analysis of designed experiments with complex aliasing. Journal of Quality Technology, 24(3):130–137, 1992.
[35] P. D. H. Hill. A review of experimental design procedures for regression model discrimination. Technometrics, 20(1):15–21, 1978.
[36] W. J. Hill, W. G. Hunter, and D. W. Wichern. A joint design criterion for the dual problem of model discrimination and parameter estimation. Technometrics, 10(1):145–160, 1968.
[37] W. G. Hunter and A. M. Reiner. Designs for discriminating between two rival models. Technometrics, 7(3):307–323, 1965.
[38] M. E. Johnson and C. J. Nachtsheim. Some guidelines for constructing exact D-optimal designs on convex design spaces. Technometrics, 25(3):271–277, 1983.
[39] B. A. Jones, W. W. Li, C. J. Nachtsheim, and Q. Y. Kenny. Model discrimination - another perspective on model-robust designs. Journal of Statistical Planning and Inference, 137(5):1576–1583, 2007.
[40] W. Ke and B. Tang. Selecting 2m-p designs using a minimum aberration criterion when some two-factor interactions are important. Technometrics, 45(4):352–360, 2003.
[41] S. Kim and W. K. Wong. Extended two-stage adaptive designs with three target responses for phase II clinical trials. Statistical Methods in Medical Research, page 0962280217709817, 2017.
[42] R.-B. Lee. Optimal model discrimination designs by discrete particle swarm optimization. Master’s thesis, National Cheng Kung University, 2015.
[43] R. Lekivetz. A new algorithm for obtaining mixed-level orthogonal and nearlyorthogonal arrays. PhD thesis, Simon Fraser University, 2006.
[44] R. Lekivetz and B. Tang. Multi-level orthogonal arrays for estimating main effects and specified interactions. Journal of Statistical Planning and Inference, 144:123–132, 2014.
[45] W. W. Li and C. J. Nachtsheim. Model-robust factorial designs. Technometrics, 42(4):345–352, 2000.
[46] W. W. Li and C. F. J. Wu. Columnwise-pairwise algorithms with applications to the construction of supersaturated designs. Technometrics, 39(2):171–179, 1997.
[47] J. L. Loeppky, R. R. Sitter, and B. Tang. Nonregular designs with desirable projection properties. Technometrics, 49(4):454–467, 2007.
[48] J. López-Fidalgo, C. Tommasi, and P. C. Trandafir. An optimal experimental design criterion for discriminating between non-normal models. Journal of the Royal Statistical Society: Series B (Statistical Methodology), 69(2):231–242, 2007.
[49] A. Mandal, W. K. Wong, and Y. Yu. Algorithmic searches for optimal designs. In A. Dean, M. Morris, J. Stufken, and D. Bingham, editors, Handbook of Design and Analysis of Experiments, pages 755–784. Chapman and Hall/CRC, Boca Raton, 2015.
[50] E. Masoudi, H. Holling, and W. K. Wong. Application of imperialist competitive algorithm to find minimax and standardized maximin optimal designs. Computational Statistics & Data Analysis, 2016.
[51] R. K. Meyer and C. J. Nachtsheim. The coordinate-exchange algorithm for constructing exact optimal experimental designs. Technometrics, 37(1):60–69, 1995.
[52] D. C. Montgomery. Design and analysis of experiments. John Wiley & Sons, 2017.
[53] M. G. Neubauer and W. Watkins. D-optimal matrices. In L. Hogben, editor, Handbook of Linear Algebra, pages 32–1–32–13. Chapman and Hall/CRC, Boca Raton, 2007.
[54] J. Nocedal and S. Wright. Numerical Optimization. Springer, 2006.
[55] F. K. H. Phoa, R.-B. Chen, W. Wang, and W. K. Wong. Optimizing two-level supersaturated designs using swarm intelligence techniques. Technometrics, 58(1):43–49, 2016.
[56] A. H. Piersma, A. Verhoef, J. te Biesebeek, M. N. Pieters, and W. Slob. Developmental toxicity of butyl benzyl phthalate in the rat using a multiple dose study design. Reproductive Toxicology, 14(5):417–425, 2000.
[57] J. Qiu, R.-B. Chen, W. Wang, and W. K. Wong. Using animal instincts to design efficient biomedical studies via particle swarm optimization. Swarm and Evolutionary Computation, 18:1–10, 2014.
[58] M. Rodriguez, B. Jones, C. M. Borror, and D. C. Montgomery. Generating and assessing exact G-optimal designs. Journal of Quality Technology, 42(1):3–20, 2010.
[59] F. J. Ruiz-Rodriguez, M. Gómez-González, and F. Jurado. Optimization of radial systems with biomass fueled gas engine from a metaheuristic and probabilistic point of view. Energy Conversion and Management, 65:343–350, 2013.
[60] N. J. A. Sloane. A library of hadamard matrices. 2012. [Online; accessed 04-July-2014].
[61] W. Slob. Dose-response modeling of continuous endpoints. Toxicological Sciences, 66(2):298–312, 2002.
[62] J. N. Srivastava. Designs for searching non-negligible effects. In J. N. Srivastava, editor, A Survey of Statistical Design and Linear Models, pages 507–519. North-Holland Publishing Co., 1975.
[63] D. X. Sun. Estimation capacity and related topics in experimental designs. PhD thesis, University of Waterloo, 1994.
[64] B. Tang and J. Zhou. Existence and construction of two-level orthogonal arrays for estimating main effects and some specified two-factor interactions. Statistica Sinica, 19(3):1193–1201, 2009.
[65] B. Tang and J. Zhou. D-optimal two-level orthogonal arrays for estimating main effects and some specified two-factor interactions. Metrika, 76(3):325–337, 2013.
[66] C. Tommasi. Optimal designs for discriminating among several non-normal models. In mODa 8-Advances in Model-Oriented Design and Analysis, pages 213–220. Springer, 2007.
[67] C. Tommasi, R. Martín-Martín, and J. López-Fidalgo. Max–min optimal discriminating designs for several statistical models. Statistics and Computing, 26(6):1163–1172, 2016.
[68] W. K. Wong, R.-B. Chen, C.-C. Huang, and W. Wang. A modified particle swarm optimization technique for finding optimal designs for mixture models. PloS one, 10(6):e0124720, 2015.
[69] C. F. J. Wu and Y. Chen. A graph-aided method for planning two-level experiments when certain interactions are important. Technometrics, 34(2):162–175, 1992.
[70] H. P. Wynn. Results in the theory and construction of D-optimum experimental designs. Journal of the Royal Statistical Society. Series B (Methodological), 34(2):133–147, 1972.
[71] H. Xu. An algorithm for constructing orthogonal and nearly-orthogonal arrays with mixed levels and small runs. Technometrics, 44(4):356–368, 2002.
[72] X.-S. Yang. Engineering Optimization: an Introduction with Metaheuristic Applications. John Wiley & Sons, 2010.
[73] S. Zhang, C. K. M. Lee, H. K. Chan, K. L. Choy, and Z. Wu. Swarm intelligence applied in green logistics: A literature review. Engineering Applications of Artificial Intelligence, 37:154–169, 2015.