狄氏分配(Dirichlet distribution)可視為高維度的貝他分配，其應用範圍包括貝氏分析的共軛先驗分配，多變量資料建模。當狄氏分配參數α_1=⋯=α_(n+1)=1時，可視為在n維空間的單體(simplex)均勻分配。高維度空間的不規則區域均勻分配有很多的應用，例如：在不規則區域中物種調查的方區抽樣和蒙地卡羅模擬(Monte Carlo Simulation)常需要多面體的均勻亂數，利用狄氏分配可更迅速的生成不規則區域的均勻亂數。本論文主要是評估由Cheng et al. (2012) 設計的R統計軟體套件“rBeta2009” [8]，並探討如何利用此套件中的狄氏分配演算法來生成其他多變量分配，如：(i)反狄氏分配(Inverted Dirichlet distribution) (ii) Liouville分配，以及(iii)由線性限制式所圍成的多面體空間之均勻分配。本文也利用電腦模擬的方式驗證本文介紹之方法比現有的電腦軟體中的演算法有效率(以電腦執行時間來看)。

Dirichlet distributions can be taken as a high-dimensioned version of beta distributions, and it has many applications, such as conjugate prior distribution in bayesian Inference and construction of the model of multivariate data. When the parameters of Dirichlet distributions are α_1=⋯=α_(n+1)=1, it can be regarded as uniform distribution within a n-dimensioned simplex. High-dimensioned uniform distribution in irregular domains has various applications, such as species surveys in quadrats sampling and Monte Carlo simulation, which often need to generate uniform random vectors over polyhedrons. With Dirichlet distributions, it is more efficient to generate uniform random vectors in irregular domain. This article evaluated the module in R, “rBeta2009” [8], originally designed by Cheng et al. (2012), and discusses how to generate other multivariate distributions by using the Dirichlet algorithm in the package, including generation of (i) Inverted Dirichlet random vectors (ii) Liouville random vectors, and (iii) uniform random vectors over polyhedrons with linear constraints. The article also verified that the method is more efficient than the older package in R. (by comparing the CPU time.)

[1] T. Bdiri and N. Bouguila (2011). Learning Inverted Dirichlet Mixtures for Positive Data Clustering. Rough Sets, Fuzzy Sets, Data Mining and Granular Computing in proc. 13th International Conference, pp. 265–272. Springer, Heidelberg.
[2] T. Bdiri and N. Bouguila (2012). Bayesian learning of inverted Dirichlet mixtures for SVM kernels generation. Neural Computing and Applications Springer-Verlag.
[3] G. Bell and M.J. Lechowicz (1991). The Ecology and Genetics of Fitness in Forest Plants. I. Environmental Heterogeneity Measured by Explant Trials. Journal of Ecology, 79(3), pp.663-685.
[4] C.P.D. Birch, S.P. Oom and J.A. Beecham (2007). Rectangular and hexagonal grids used for observation, experiment and simulation in ecolo-gy.Ecological Modelling, 206(3), pp.347-359.
[5] J.H. Blanche and L. Rojas-nandayapa (2011). Efficient Simulation of Tail Probabilities of Sums of Dependent Random Variables. Applied Probability Trust, 48A, pp. 147-164.
[6] K.G. van den Boogaart, R. Tolosana and M. Bren (2013). Compositional Data Analysis in R : the Package “compositions”.
[7] N. Bouguila (2011). A Liouville-Based Approach for Discrete Data Categorization.Rough Sets, Fuzzy Sets, Data Mining and Granular Computing in proc. 13th International Conference, pp 330-337, Russia, Springer-Verlag.
[8] C.W. Cheng, Y.C. Hung and N. Balakrishnan (2012) Generating Beta Random Numbers and Dirichlet Random Vectors in R : the Package “rBeta2009”, Accepted by Computational Statistics & Data Analysis.
[9] C. Elkan (2006). Clustering documents with an exponential-family ap-proximation of the Dirichlet compound multinomial distribution in Proc. the 23rd international conf. Machine learning, N.Y. USA, pp.289-296.
[10] K.T. Fang, G.L. Tian and M.Y. Xie. (1997). Uniform Distribution on Convex Polyhedron and Its Applications. Technical report No. 149, Depart-ment of Mathematics, Hong Kong Batist University.
[11] R.D. Gupta and D. Richards (1987). Multivariate Liouville distributions Journal of Multivariate Analysis, 23, pp. 233–256.
[12] R.D. Gupta and D. St.P Richards (1990). The Dirichlet distributions and polynomial regression. Journal of Multivariate Analysis, 32(1), pp.95-102.
[13] R.D. Gupta and D. St. P. Richards (2001). The History of the Dirichlet and Liouville Distributions. International Statistical Review, 69(3), pp.433-446.
[14] I. Guttman and G.C. Tiao (1965). The inverted Dirichlet distribution with applications. Journal of American Statistics Association, 60, pp. 793–805.
[15] Y.C. Hung, N. Balakrishnan and C.W. Cheng (2011). Evaluation of algorithms for generating Dirichlet random vectors. Journal of Statistical Computation and Simulation, 81(4), p.445-459.
[16] Y.C. Hung, N. Balakrishnan and Y.T. Lin (2009). Evaluation of Beta Generation Algorithms. Communications in Statistics - Simulation and Com-putation, 38, pp. 750-770.
[17] D.P. Kennedy (1988). A note on stochastic search methods for global optimization. Advances in Applied Probability, 20, pp. 476-478.
[18] K. Lange (1995). Applications of the Dirichlet distribution to forensic match probabilities. Genetica, 96, pp.107-117.
[19] J. Liouville (1839). Note sur quelquess integrals définies. Journal deMa-thématiques Pures et Appliquées, 4, pp. 225-235.
[20] R.E. Madsen, D. Kauchak and C. Elkan (2005). Modeling Word Bursti-ness Using the Dirichlet Distribution. Proceeding of the 22nd International Conference on Machine Learning, pp. 545-552.
[21] A. Marshall and I. Olkin (1979). Inequalities: Theory of Majorization and Its Applications. New York Academic Press, pp.416-417.
[22]B. E. McNeil, R. E. Martell and J. M. Read (2006). GIS and biogeochemical models for examining the legacy of forest disturbance in the Adirondack Park, NY, USA. Ecological Modelling, 195(3), pp.281-295.
[23] A. J. McNeila and J. Nešlehová (2010). From Archimedean to Liouville copulas. Journal of Multivariate Analysis, 101(8), pp. 1772-1790.
[24] A. D. Martin, K. M. Quinn and J. H. Park (2012). Markov chain Monte Carlo (MCMC) in R : the Package “MCMCpack”.
[25] A. Narayanan (1990). Computer generation of Dirichlet random vectors. Journal of Statistical Computation and Simulation, 36, pp. 19–30.
[26] K.W. Ng, M.L. Yang, M. Tan and G.L. Tian (2008). Grouped Dirichlet distribution : A new tool for incomplete categorical data analysis. Journal of Multivariate Analysis, 99, pp. 490-509.
[27] D.G. Rameshwar and St. P.R. Donald (1987). Multivariate Liouville dis-tribution. Journal of Multivariate Analysis, 23, pp. 233-256.
[28] R.Y. Rubinstein (1982). Generating random vectors uniformly distributed inside and on the surface of different regions. European Journal of Operational Research, 10(2), pg. 205-209.
[29] H. Sahai and R.L. Anderson (1973). Confidence regions for variance ratios of the random models for balanced data. Journal of the American Statistical Association, 68(344), pp. 951-952.
[30] H. Sakasegawa (1983). Stratified rejection and squeeze method for generating beta random numbers. Annals of the Institute Statistical Mathematics, 35, pp. 291-302.
[31] B.W. Schmeiser and A.J.G. Babu (1980). Beta Variate Generation via Exponential Majorizing Functions. Operations Research, 28, pp. 917-926.
[32] R.L. Smith (1984) Efficient Monte Carlo Procedures for Generating Points Uniformly Distributed over Bounded Regions. Operations search, 32(6), pp. 1296-1308.
[33] D. Stubbs, A. Hailey, H. Pulford and W. Tyler (1984). Population Ecology of European Tortoises: Review of Field Techniques. Amphibia-Reptilia, 5(1), pp. 57-68(12).
[34] G.L. Tian, M.L. Tang, K.C. Yuen and K.W. Ng (2010). Further properties and new applications of the nested Dirichlet distribution.Computational Statistics And Data Analysis, 5(2), p. 394-405.

[35] G.G. Tiao and I. Guttman (1965). The inverted Dirichlet distribution with applications. Journal of the American Statistical Association, 60, pp. 793–805.
[36] G.G. Tiao and I. Guttman (1965). The inverted Dirichlet distribution with applications. Journal of the American Statistical Association, 60, pp. 1251-1252.
[37] T.T. Wong (1998). Generalized Dirichlet distribution in Bayesian analysis. Applied Mathematics and Computation, 97, pp.165-181.
[38] H. Yassaee (1974) Inverted dirichlet distribution and multivariate logistic distribution.Canadian Journal of Statistics, 2, pp.99-105.
[39] H. Zechner and E. Stadlober (1993). Generating beta variates via patch-work rejection. Computing, 50, pp. 1-18.