When lattice bases are Markov bases. (English) Zbl 07878517
Summary: Sampling the fibre comprising the solutions of a linear inverse problem for count data is an important practical problem. Connectivity of the sampler is guaranteed only if a Markov basis, defining a sufficiently rich variety of sampling directions, is available. Computation of a Markov basis is typically challenging, and the mixing properties of the resulting sampler can be poor. However, for some problems a suitably chosen lattice basis will be a Markov basis. We provide an easily checkable condition for the existence of such a lattice Markov basis, and demonstrate that associated hit-and-run samplers will mix rapidly for uniform target distributions.
MSC:
| 62P10 | Applications of statistics to biology and medical sciences; meta analysis |
| 62F15 | Bayesian inference |
| 62M99 | Inference from stochastic processes |
References:
| [1] | Airoldi, E., Haas, B., 2011. Polytope samplers for inference in ill-posed inverse problems. In: International Conference on Artificial Intelligence and Statistics, Vol. 15. pp. 110-118. |
| [2] | Aoki, S.; Hara, H.; Takemura, A., Markov Bases in Algebraic Statistics, 2012, Springer: Springer New York · Zbl 1304.62015 |
| [3] | Diaconis, P.; Sturmfels, B., Algebraic algorithms for sampling from conditional distributions, Ann. Statist., 26, 1, 363-397, 1998 · Zbl 0952.62088 |
| [4] | Dobra, A.; Karr, A. F.; Sanil, A. P., Preserving confidentiality of high-dimensional tabulated data: statistical and computational issues, Stat. Comput., 13, 4, 363-370, 2003 |
| [5] | Fischer, K.; Morris, W.; Shapiro, J., Mixed dominating matrices, Linear Algebra Appl., 270, 1-3, 191-214, 1998 · Zbl 0889.15020 |
| [6] | Fischer, K.; Shapiro, J., Mixed matrices and binomial ideals, J. Pure Appl. Algebra, 113, 1, 39-54, 1996 · Zbl 0864.15016 |
| [7] | Hazelton, M. L., Estimation of origin-destination trip rates in leicester, J. R. Stat. Soc. Ser. C, 50, 4, 423-433, 2001 · Zbl 1112.62318 |
| [8] | Hazelton, M. L., Statistical inference for transit system origin-destination matrices, Technometrics, 52, 2, 221-230, 2010 |
| [9] | Hazelton, M. L., Network tomography for integer-valued traffic, Ann. Appl. Stat., 9, 1, 474-506, 2015 · Zbl 1454.62515 |
| [10] | Hazelton, M.; McVeagh, M.; van Brunt, B., Geometrically aware dynamic Markov bases for statistical linear inverse problems, Biometrika, 108, 3, 609-626, 2021 · Zbl 07459719 |
| [11] | Hazelton, M.; McVeagh, M.; van Brunt, B., Some rapidly mixing hit-and-run samplers for latent counts in linear inverse problems, Bernoulli, 2024, (in press) · Zbl 07898695 |
| [12] | Levin, D.; Peres, Y.; Wilmer, E., Markov Chains and Mixing Times, 2009, American Mathematical Soc · Zbl 1160.60001 |
| [13] | Schofield, M. R.; Bonner, S. J., Connecting the latent multinomial, Biometrics, 71, 1070-1080, 2015 · Zbl 1419.62436 |
| [14] | Smith, R., Efficient Monte Carlo procedures for generating points uniformly distributed over bounded regions, Oper. Res., 32, 6, 1296-1308, 1984 · Zbl 0552.65004 |
| [15] | Stanley, C.; Windisch, T., Heat-bath random walks with Markov bases, Adv. Appl. Math., 92, 122-143, 2018 · Zbl 1371.05276 |
| [16] | Sutherland, J.; Schwarz, C., Multi-list methods using incomplete lists in closed populations, Biometrics, 61, 1, 134-140, 2005 · Zbl 1077.62115 |
| [17] | Tebaldi, C.; West, M., Bayesian inference on network traffic using link count data (with discussion), J. Amer. Statist. Assoc., 93, 557, 557-576, 1998 · Zbl 1072.62650 |
This reference list is based on information provided by the publisher or from digital mathematics libraries. Its items are heuristically matched to zbMATH identifiers and may contain data conversion errors. In some cases that data have been complemented/enhanced by data from zbMATH Open. This attempts to reflect the references listed in the original paper as accurately as possible without claiming completeness or a perfect matching.
