## The Annals of Statistics

### Approximating faces of marginal polytopes in discrete hierarchical models

#### Abstract

The existence of the maximum likelihood estimate in a hierarchical log-linear model is crucial to the reliability of inference for this model. Determining whether the estimate exists is equivalent to finding whether the sufficient statistics vector $t$ belongs to the boundary of the marginal polytope of the model. The dimension of the smallest face $\mathbf{F}_{t}$ containing $t$ determines the dimension of the reduced model which should be considered for correct inference. For higher-dimensional problems, it is not possible to compute $\mathbf{F}_{t}$ exactly. Massam and Wang (2015) found an outer approximation to $\mathbf{F}_{t}$ using a collection of submodels of the original model. This paper refines the methodology to find an outer approximation and devises a new methodology to find an inner approximation. The inner approximation is given not in terms of a face of the marginal polytope, but in terms of a subset of the vertices of $\mathbf{F}_{t}$.

Knowing $\mathbf{F}_{t}$ exactly indicates which cell probabilities have maximum likelihood estimates equal to $0$. When $\mathbf{F}_{t}$ cannot be obtained exactly, we can use, first, the outer approximation $\mathbf{F}_{2}$ to reduce the dimension of the problem and then the inner approximation $\mathbf{F}_{1}$ to obtain correct estimates of cell probabilities corresponding to elements of $\mathbf{F}_{1}$ and improve the estimates of the remaining probabilities corresponding to elements in $\mathbf{F}_{2}\setminus\mathbf{F}_{1}$. Using both real-world and simulated data, we illustrate our results, and show that our methodology scales to high dimensions.

#### Article information

Source
Ann. Statist., Volume 47, Number 3 (2019), 1203-1233.

Dates
Revised: April 2018
First available in Project Euclid: 13 February 2019

https://projecteuclid.org/euclid.aos/1550026834

Digital Object Identifier
doi:10.1214/18-AOS1710

Mathematical Reviews number (MathSciNet)
MR3911110

Zentralblatt MATH identifier
07053506

#### Citation

Wang, Nanwei; Rauh, Johannes; Massam, Hélène. Approximating faces of marginal polytopes in discrete hierarchical models. Ann. Statist. 47 (2019), no. 3, 1203--1233. doi:10.1214/18-AOS1710. https://projecteuclid.org/euclid.aos/1550026834

#### References

• Banerjee, O., El Ghaoui, L. and d’Aspremont, A. (2008). Model selection through sparse maximum likelihood estimation for multivariate Gaussian or binary data. J. Mach. Learn. Res. 9 485–516.
• Barndorff-Nielsen, O. (1978). Information and Exponential Families in Statistical Theory. Wiley, Chichester.
• Csiszár, I. and Matúš, F. (2008). Generalized maximum likelihood estimates for exponential families. Probab. Theory Related Fields 141 213–246.
• Csiszár, I. and Shields, P. (2004). Information Theory and Statistics: A Tutorial, 1st ed. Now Publishers, Hanover, MA.
• Deza, M. M. and Laurent, M. (2010). Geometry of Cuts and Metrics. Algorithms and Combinatorics 15. Springer, Heidelberg.
• Dobra, A., Erosheva, E. A. and Fienberg, S. E. (2004). Disclosure limitation methods based on bounds for large contingency tables with applications to disability. In Statistical Data Mining and Knowledge Discovery 93–116. Chapman & Hall, Boca Raton, FL.
• Dobra, A. and Lenkoski, A. (2011). Copula Gaussian graphical models and their application to modeling functional disability data. Ann. Appl. Stat. 5 969–993.
• Eriksson, N., Fienberg, S. E., Rinaldo, A. and Sullivant, S. (2006). Polyhedral conditions for the nonexistence of the MLE for hierarchical log-linear models. J. Symbolic Comput. 41 222–233.
• Fienberg, S. E. (1980). The Analysis of Cross-Classified Categorical Data, 2nd ed. MIT Press, Cambridge, MA.
• Fienberg, S. E. and Rinaldo, A. (2007). Three centuries of categorical data analysis: Log-linear models and maximum likelihood estimation. J. Statist. Plann. Inference 137 3430–3445.
• Fienberg, S. E. and Rinaldo, A. (2012). Maximum likelihood estimation in log-linear models. Ann. Statist. 40 996–1023.
• Gawrilow, E. and Joswig, M. (2000). Polymake: A framework for analyzing convex polytopes. In Polytopes—Combinatorics and Computation (Oberwolfach, 1997). DMV Sem. 29 43–73. Birkhäuser, Basel.
• Geyer, C. J. (2009). Likelihood inference in exponential families and directions of recession. Electron. J. Stat. 3 259–289.
• Haberman, S. J. (1974). The Analysis of Frequency Data. The Univ. Chicago Press, Chicago, IL.
• Karwa, V. and Slavković, A. (2016). Inference using noisy degrees: Differentially private $\beta$-model and synthetic graphs. Ann. Statist. 44 87–112.
• Lauritzen, S. L. (1996). Graphical Models. Oxford Statistical Science Series 17. Oxford Univ. Press, New York.
• Letac, G. and Massam, H. (2012). Bayes factors and the geometry of discrete hierarchical loglinear models. Ann. Statist. 40 861–890.
• Liu, Q. and Ihler, A. (2012). Distributed parameter estimation via pseudo-likelihood. Int. Conf. Mach. Learn. (ICML).
• Massam, H. and Wang, N. (2015). A local approach to estimation in discrete loglinear models. Preprint. Available at arXiv:1504.05434.
• Massam, H. and Wang, N. (2018). Local conditional and marginal approach to parameter estimation in discrete graphical models. J. Multivariate Anal. 164 1–21.
• Rauh, J., Kahle, T. and Ay, N. (2011). Support sets in exponential families and oriented matroid theory. Internat. J. Approx. Reason. 52 613–626.
• Ravikumar, P., Wainwright, M. J. and Lafferty, J. D. (2010). High-dimensional Ising model selection using $\ell_{1}$-regularized logistic regression. Ann. Statist. 38 1287–1319.
• Schmidt, M. (2005). minFunc: Unconstrained differentiable multivariate optimization in Matlab. http://www.cs.ubc.ca/~schmidtm/Software/minFunc.html.
• Vlach, M. (1986). Conditions for the existence of solutions of the three-dimensional planar transportation problem. Discrete Appl. Math. 13 61–78.
• Wang, N., Rauh J. and Massam, H. (2019). Supplement to “Approximating faces of marginal polytopes in discrete hierarchical models.” DOI:10.1214/18-AOS1710SUPP.
• Ziegler, G. M. (1995). Lectures on Polytopes. Graduate Texts in Mathematics 152. Springer, New York.

#### Supplemental materials

• Supplement to “Approximating faces of marginal polytopes in discrete hierarchical models.”. Appendix A describes the concrete parametrization that we use in the examples. Appendix B discusses the case of two binary variables to illustrate what happens to the usual parameters when the MLE does not exist. Appendix C discusses how to further improve the parametrization $\mu_{L}$ introduced in Section 2. Appendices D and E give further results for the examples from Section 5. Appendix F gives the technical details for the example in Section 6.2.