Published on Tue Dec 30 2014

On Semiparametric Exponential Family Graphical Models

Zhuoran Yang, Yang Ning, Han Liu

We propose a new class of semiparametric exponential family graphical models for analysis of high dimensional mixed data. We consider both problems of parameter estimation and hypothesis testing in high dimensions. We propose a symmetric pairwise score test for the presence of a single edge in the graph.

0
0
0
Abstract

We propose a new class of semiparametric exponential family graphical models for the analysis of high dimensional mixed data. Different from the existing mixed graphical models, we allow the nodewise conditional distributions to be semiparametric generalized linear models with unspecified base measure functions. Thus, one advantage of our method is that it is unnecessary to specify the type of each node and the method is more convenient to apply in practice. Under the proposed model, we consider both problems of parameter estimation and hypothesis testing in high dimensions. In particular, we propose a symmetric pairwise score test for the presence of a single edge in the graph. Compared to the existing methods for hypothesis tests, our approach takes into account of the symmetry of the parameters, such that the inferential results are invariant with respect to the different parametrizations of the same edge. Thorough numerical simulations and a real data example are provided to back up our results.

Thu Feb 26 2015
Machine Learning
ROCKET: Robust Confidence Intervals via Kendall's Tau for Transelliptical Graphical Models
The transelliptical model generalizes Gaussian and nonparanormal models to a broader family of distributions. We propose the ROCKET method, which constructs an estimator of that we prove to be asymptotically normal under mild assumptions.
0
0
0
Mon Dec 03 2012
Machine Learning
Structure estimation for discrete graphical models: Generalized covariance matrices and their inverses
For certain graph structures, the support of the inverse. of indicator variables on the vertices of a graph reflects the.conditional independence structure of the graph. These population-level results have implications for graph selection methods, both known and novel.
0
0
0
Fri Jun 08 2012
Machine Learning
Multivariate Bernoulli distribution
In this paper, we consider the multivariate Bernoulli distribution as a model to estimate the structure of graphs with binary nodes. This distribution is discussed in the framework of the exponential family, and its statistical properties regarding independence of the nodes are demonstrated.
0
0
0
Tue Apr 21 2015
Machine Learning
A local approach to estimation in discrete loglinear models
When the data is sparse, there are many zeros in the contingency table and the maximum likelihood estimate of the parameter may not exist. Fienberg and Rinaldo have shown that the mle does not exists iff the data vector does not belong to a face of the so-called marginal cone spanned by the rows of the design matrix.
0
0
0
Mon Oct 19 2020
Machine Learning
Learning Exponential Family Graphical Models with Latent Variables using Regularized Conditional Likelihood
Fitting a graphical model to a collection of random variables is a challenging task if the observed variables are influenced by latent variables. We present a new convex relaxation framework based on regularized conditional likelihood.
0
0
0
Mon Oct 21 2013
Machine Learning
Distributed parameter estimation of discrete hierarchical models via marginal likelihoods
We consider discrete graphical models Markov with respect to a graph and propose two distributed marginal methods. Both methods are based on arelaxation of the marginal likelihood. The two methods differ by the size of the neighborhood of .
0
0
0