Electronic Journal of Statistics

On the geometry of discrete exponential families with application to exponential random graph models

Alessandro Rinaldo, Stephen E. Fienberg, and Yi Zhou

Full-text: Open access

Abstract

There has been an explosion of interest in statistical models for analyzing network data, and considerable interest in the class of exponential random graph (ERG) models, especially in connection with difficulties in computing maximum likelihood estimates. The issues associated with these difficulties relate to the broader structure of discrete exponential families. This paper re-examines the issues in two parts. First we consider the closure of k-dimensional exponential families of distribution with discrete base measure and polyhedral convex support P. We show that the normal fan of P is a geometric object that plays a fundamental role in deriving the statistical and geometric properties of the corresponding extended exponential families. We discuss its relevance to maximum likelihood estimation, both from a theoretical and computational standpoint. Second, we apply our results to the analysis of ERG models. By means of a detailed example, we provide some characterization of the properties of ERG models, and, in particular, of certain behaviors of ERG models known as degeneracy.

Article information

Source
Electron. J. Statist. Volume 3 (2009), 446-484.

Dates
First available: 26 May 2009

Permanent link to this document
http://projecteuclid.org/euclid.ejs/1243343761

Digital Object Identifier
doi:10.1214/08-EJS350

Mathematical Reviews number (MathSciNet)
MR2507456

Zentralblatt MATH identifier
06166450

Subjects
Primary: 62F99: None of the above, but in this section
Secondary: 62F99: None of the above, but in this section

Keywords
Exponential families normal cone maximum likelihood estimates exponential random graphs

Citation

Rinaldo, Alessandro; Fienberg, Stephen E.; Zhou, Yi. On the geometry of discrete exponential families with application to exponential random graph models. Electronic Journal of Statistics 3 (2009), 446--484. doi:10.1214/08-EJS350. http://projecteuclid.org/euclid.ejs/1243343761.


Export citation

References

  • Bardorff-Nielsen, O. (1978)., Information and Exponential Families in Statistical Theory, New York: John Wiley & Sons, New York.
  • Bardorff-Nielsen, O. (2006). Koopman–Darmois–Pitman Families, Encycploedia of Statistical Sciences.
  • Bishop, Y.M.M., Fienberg, S.E., and Holland, P.W. (1975)., Discrete Multivariate Analysis: Theory and Practice. MIT Press, Cambridge, MA, 1975. Reprinted by Springer-Verlag, New York, 2007.
  • Brown, L. (1986)., Fundamentals of Statistical Exponential Families, IMS Lecture Notes-Monograph Series, Vol.9, Hayward, CA.
  • Csiszár, I. and Matúš, F. (2008). Generalized maximum likelihood estimates for exponential families, Probability Theory and Related Fields, 141, 213–246.
  • Csiszár, I. and Matúš, F. (2005). Closure of exponential families, Annals of Probability, 33 (2), 582–600.
  • Csiszár, I. and Matúš, F. (2003). Information projection revisited, IEEE Transaction of Information Theory, 49 (6), 1474–1490.
  • Csiszár, I. and Matúš, F. (2001). Convex cores of measures, Studia Scientiarum Mathematicarum Hungarica, 38, 177–190.
  • Eriksson, N., Fienberg, S.E., Rinaldo, A. and Sullivant S. (2006). Polyhedral conditions for the nonexistence of the MLE for hierarchical log-linear models, Journal of Symbolic Computation, 41, 222–233.
  • Fienberg, S.E. and Wasserman, S. (1981). An exponential family of probability distributions for directed graphs: comment, Journal of the American Statistical Association, 76(373), 54–57.
  • Frank, O. and Strauss, D. (1986). Markov Graphs, Journal of the American Statistical Association, 81, 832–842.
  • Geiger, A., Meek, C. and Sturmfels, B. (2006). On the toric algebra of graphical models, Annals of Statistics, 34(3), 1463–1492.
  • Geyer, C. (2008). Likelihood Inference in Exponential Families and Directions of Recession, Technical Report TR 672, Department of Statistics, University of, Minnesota.
  • Geyer, C.J. and Thompson, E.A. (1992). Constrained Monte Carlo maximum likelihood calculations (with discussion), Journal of the Royal Statistical Society, Series B, 54, 657–699.
  • Goldenberg, A., Zheng, A.X., Fienberg, S.E., and Airoldi, E.M. (2009). A survey of statistical network models, Foundations and Trends in Machine Learning, to appear.
  • Goodreau, S. (2007). Advances in exponential random graph, (p*) models applied to a large social network, Social Networks, 29, 231–248.
  • Haberman, S.J. (1981). An exponential family of probability distributions for directed graphs: comment, Journal of the American Statistical Association, 76(373), 60–61.
  • Handcock, M.S., Hunter, D., Butts, C., Goodreau, S. and Morris, M. (2006). Statnet: An R Package for the Statistical Analysis and Simulation of Social Networks. Manual. University of Washington, http://www.csde., washington.edu/statnet.
  • Handcock, M.S. (2003). Assessing degeneracy in statistical models for social networks. Working paper 39, Center for Statistics and the Social Sciences, University of, Washington.
  • Holland, P.W. and Leinhardt, S. (1981). An exponential family of probability distributions for directed graphs (with discussion)., Journal of the American Statistical Association 76 (373), 33–65.
  • Hunter, D.R., Goodreau, S.M. and Handcock, M.S. (2008). Goodness of fit of social network models, Journal of the American Statistical Association, 103(481), 248–258.
  • Lee, Y. and Nelder, J.A. (1996). Hierarchical generalized linear models (with discussion), Journal of the Royal Statistical Society, Series B, 58, 619–678.
  • Letac, G. (1992)., Lectures on Natural Exponential Families and Their Variance Functions, Monografias de Matemática 50. Instituto de Matemática Pura e Aplicada, Rio de Janeiro, Brazil.
  • Moreno, J.L. (1934)., Who Shall Survive? Nervous and Mental Disease Publishing Company, Washington, DC.
  • Rinaldo, A. (2006a). On maximum likelihood estimation in log-linear models, Technical Report 833, Department of Statistics, Carnegie Mellon, University.
  • Rinaldo, A. (2006b). Computing Maximum Likelihood Estimates in Log-Linear Models, Technical Report 835, Department of Statistics, Carnegie Mellon, University.
  • Robins, G., Pattison, P., Kalish, Y. and Lusher, D. (2007). An introduction to exponential random graph, (p*) models for social networks, Social Networks, 29, 171–191.
  • Robins, G, Snijers, T., Wang, P., Handcock, M. and Pattison, P. (2007). Recent developments in exponential random graph, (p*) models for social networks, Social Networks, 29, 192–215.
  • Rockafellar, R.T. (1970)., Convex Analysis, Princeton University Press, Princeton.
  • Schrijver, A. (1998)., Theory of Linear and Integer Programming, Wiley & Sons, New York.
  • Snijders, T.A.B. (2002). Markov Chain Monte Carlo estimation of exponential random graph models, Journal of Social Structure, 3.
  • Strauss, D. and M. Ikeda (1990). Pseudolikelihood estimation for social networks, Journal of the American Statistical Association, 85, 204–212.
  • Sturmfels, B. (1995)., Gröbner Bases and Convex Polytopes, American Mathematical Society, Providence, RI.
  • Wasserman, S. and Pattison, P.E. (1996). Logit models and logistic regressions for social networks: I. An introduction to Markov graphs and, p*, Psychometrika 61, 401–425.
  • Wasserman, S.S. and Robins, G.L. (2004). An Introduction to Random Graphs, Dependence Graphs, and, p*, in Models and Methods in Social Network Analysis, eds. P. Carrington, J.S. and Wasserman, S.S., Cambridge: Cambridge University Press, New York.
  • Ziegler, G.M. (2001). Lectures on 0/1 polytopes, in Polytopes—Combinatorics and Computations, DMS Seminar, Band 29, G. Kalai and G.M. Ziegler editors, Birkhäuser.
  • Ziegler, G.M. (1995)., Lectures on Polytopes, Springer-Verlag, New York.