Canonical representation of conditionally specified multivariate discrete distributions

Most work on conditionally specified distributions has focused on approaches that operate on the probability space, and the constraints on the probability space often make the study of their properties challenging. We propose decomposing both the joint and conditional discrete distributions into characterizing sets of canonical interactions, and we prove that certain interactions of a joint distribution are shared with its conditional distributions. This invariance opens the door for checking the compatibility between conditional distributions involving the same set of variables. We formulate necessary and sufficient conditions for the existence and uniqueness of discrete conditional models, and we show how a joint distribution can be easily computed from the pool of interactions collected from the conditional distributions. Hence, the methods can be used to calculate the exact distribution of a Gibbs sampler. Furthermore, issues such as how near compatibility can be reconciled are also discussed. Using mixed parametrization, we show that the proposed approach is based on the canonical parameters, while the conventional approaches are based on the mean parameters. Our advantage is partly due to the invariance that holds only for the canonical parameters.

[1]  Donald B. Rubin,et al.  Nested multiple imputation of NMES via partially incompatible MCMC , 2003 .

[2]  T. Speed,et al.  Characterizing a joint probability distribution by conditionals , 1993 .

[3]  B. Arnold,et al.  Compatibility of partial or complete conditional probability specifications , 2004 .

[4]  George Casella,et al.  Functional Compatibility, Markov Chains and Gibbs Sampling with Improper Posteriors , 1998 .

[5]  B. Arnold,et al.  Conditionally Specified Distributions: An Introduction (with comments and a rejoinder by the authors) , 2001 .

[6]  B. Arnold,et al.  Compatible Conditional Distributions , 1989 .

[7]  Yuchung J. Wang,et al.  A Strategy for Designing Telescoping Models for Analyzing Multiway Contingency Tables Using Mixed Parameters , 2003 .

[8]  B. Arnold,et al.  Conditional specification of statistical models , 1999 .

[9]  Edward H. Ip,et al.  Locally dependent latent trait model for polytomous responses with application to inventory of hostility , 2004 .

[10]  Enrique Castillo,et al.  Exact and near compatibility of discrete conditional distributions , 2002 .

[11]  Giuseppe Toffoli,et al.  The role of UGT1A1*28 polymorphism in the pharmacodynamics and pharmacokinetics of irinotecan in patients with metastatic colorectal cancer. , 2006, Journal of clinical oncology : official journal of the American Society of Clinical Oncology.

[12]  T. Speed,et al.  Corrigendum: Characterizing a joint probability distribution by conditionals , 1999 .

[13]  B. Arnold,et al.  Specification of distributions by combinations of marginal and conditional distributions , 1996 .

[14]  E. Barankin.  Sufficient parameters: Solution of the minimal dimensionality problem , 1960 .

[15]  Julian Besag,et al.  Discussion: Markov Chains for Exploring Posterior Distributions , 1994 .

[16]  Stef van Buuren,et al.  Multiple imputation of discrete and continuous data by fully conditional specification , 2007 .

[17]  J. F. C. Kingman,et al.  Information and Exponential Families in Statistical Theory , 1980 .

[18]  David Maxwell Chickering,et al.  Dependency Networks for Inference, Collaborative Filtering, and Data Visualization , 2000, J. Mach. Learn. Res..