Abstract
For general (i.e., non decomposable) hierarchical models of contingency tables, structure of a Markov basis is very complicated. It contains moves of higher degree in general and the complexity increases as the sizes of contingency tables increase. However, Markov chain Monte Carlo methods for non-decomposable models are very useful since the closed form expression of null distributions cannot be obtained for general non-decomposable models and therefore other methods such as exact methods cannot be used. In this chapter, we first consider the no-three-factor interaction model of three-way contingency tables as an example of non-decomposable models. The complete structure of Markov bases for this model has not yet been obtained at present. We give a structure of minimal Markov bases for 3 ×3 ×K contingency tables with fixed two-dimensional marginals. We then discuss Markov bases of reducible models, Markov complexity and Markov width.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
References
ti2 team: 4ti2 — a software package for algebraic, geometric and combinatorial problems on linear spaces. Available at www.4ti2.de
Aoki, S., Takemura, A.: The list of indispensable moves of the unique minimal markov basis for 3 ×4 ×k and 4 ×4 ×4 contingency tables with fixed two-dimensional marginals. Tech. Rep. METR 2003-38, University of Tokyo (2003)
Aoki, S., Takemura, A.: Minimal basis for a connected Markov chain over 3 ×3 ×K contingency tables with fixed two-dimensional marginals. Aust. N. Z. J. Stat. 45(2), 229–249 (2003)
Aoki, S., Takemura, A.: Statistics and Gröbner bases—the origin and development of computational algebraic statistics. In: Selected papers on probability and statistics, Amer. Math. Soc. Transl. Ser. 2, vol. 227, pp. 125–145. Amer. Math. Soc., Providence, RI (2009)
Badsberg, J.H., Malvestuto, F.M.: An implementaition of the iterative proportional fitting procecure by propagation trees. Comput. Statist. Data. Anal. 37, 297–322 (2001)
Berstein, Y., Onn, S.: The Graver complexity of integer programming. Ann. Comb. 13(3), 289–296 (2009)
Boffi, G., Rossi, F.: Lexicographic Gröbner bases for transportation problems of format r ×3 ×3. J. Symbolic Comput. 41(3-4), 336–356 (2006)
Bruns, W., Hemmecke, R., Ichim, B., Köppe, M., Söger, C.: Challenging computations of Hilbert bases of cones associated with algebraic statistics. Exp. Math. 20(1), 25–33 (2011)
De Loera, J.A., Hemmecke, R., Onn, S., Weismantel, R.: n-fold integer programming. Discrete Optim. 5(2), 231–241 (2008)
De Loera, J.A., Onn, S.: All linear and integer programs are slim 3-way transportation programs. SIAM J. Optim. 17(3), 806–821 (2006)
Develin, M., Sullivant, S.: Markov bases of binary graph models. Ann. Comb. 7(4), 441–466 (2003)
Dobra, A., Sullivant, S.: A divide-and-conquer algorithm for generating Markov bases of multi-way tables. Comput. Statist. 19(3), 347–366 (2004)
Hara, H., Sei, T., Takemura, A.: Hierarchical subspace models for contingency tables. J. Multivariate Anal. 103, 19–34 (2012)
Hara, H., Takemura, A.: A localization approach to improve iterative proportional scaling in gaussian graphical models. Comm. Stat. Theor. Meth. 39, 1643–1654 (2010)
Hemmecke, R., Malkin, P.N.: Computing generating sets of lattice ideals and Markov bases of lattices. J. Symbolic Comput. 44(10), 1463–1476 (2009)
Hemmecke, R., Nairn, K.A.: On the Gröbner complexity of matrices. J. Pure Appl. Algebra 213(8), 1558–1563 (2009)
Hoşten, S., Sullivant, S.: Gröbner bases and polyhedral geometry of reducible and cyclic models. J. Combin. Theory Ser. A 100(2), 277–301 (2002)
Hoşten, S., Sullivant, S.: A finiteness theorem for Markov bases of hierarchical models. J. Combin. Theory Ser. A 114(2), 311–321 (2007)
Lauritzen, S.L.: Graphical Models. Oxford University Press, Oxford (1996)
Leimer, H.G.: Optimal decomposition by clique separators. Discrete Math. 113, 99–123 (1993)
Malvestuto, F.M., Moscarini, M.: Decomposition of a hypergraph by partial-edge separators. Theoret. Comput. Sci. 237, 57–79 (2000)
Ohsugi, H., Hibi, T.: Toric ideals arising from contingency tables. In: Proceedings of the Ramanujan Mathematical Society’s Lecture Notes Series pp. 87–111 (2006)
Ohsugi, H., Hibi, T.: Non-very ample configurations arising from contingency tables. Ann. Inst. Statist. Math. 62(4), 639–644 (2010)
Petrović, S., Stokes, E.: Betti numbers of Stanley-Reisner rings determine hierarchical Markov degrees (2012). To appear in Journal of Algebraic Combinatorics
Santos, F., Sturmfels, B.: Higher Lawrence configurations. J. Combin. Theory Ser. A 103(1), 151–164 (2003)
Vlach, M.: Conditions for the existence of solutions of the three-dimensional planar transportation problem. Discrete Appl. Math. 13(1), 61–78 (1986)
Author information
Authors and Affiliations
Rights and permissions
Copyright information
© 2012 Springer Science+Business Media New York
About this chapter
Cite this chapter
Aoki, S., Hara, H., Takemura, A. (2012). Markov Basis for No-Three-Factor Interaction Models and Some Other Hierarchical Models. In: Markov Bases in Algebraic Statistics. Springer Series in Statistics, vol 199. Springer, New York, NY. https://doi.org/10.1007/978-1-4614-3719-2_9
Download citation
DOI: https://doi.org/10.1007/978-1-4614-3719-2_9
Published:
Publisher Name: Springer, New York, NY
Print ISBN: 978-1-4614-3718-5
Online ISBN: 978-1-4614-3719-2
eBook Packages: Mathematics and StatisticsMathematics and Statistics (R0)