Markov Chain Monte Carlo using Tree-Based Priors on Model Structure
نویسندگان
چکیده
We present a general framework for defining priors on model structure and sampling from the posterior using the Metropolis-Hastings algo rithm. The key ideas are that structure priors are defined via a probability tree and that the pro posal distribution for the Metropolis-Hastings al gorithm is defined using the prior, thereby defin ing a cheaply computable acceptance probabil ity. We have applied this approach to Bayesian net structure learning using a number of priors and proposal distributions. Our results show that these must be chosen appropriately for this ap proach to be successful.
منابع مشابه
Improving phylogenetic analyses by incorporating additional information from genetic sequence databases
MOTIVATION Statistical analyses of phylogenetic data culminate in uncertain estimates of underlying model parameters. Lack of additional data hinders the ability to reduce this uncertainty, as the original phylogenetic dataset is often complete, containing the entire gene or genome information available for the given set of taxa. Informative priors in a Bayesian analysis can reduce posterior un...
متن کاملBayesian analysis of software reliability models with reference prior
In this paper, we introduce a Bayesian analysis for non-homogeneous Poisson process in software reliability models. Posterior summaries of interest are obtained using Markov chain Monte Carlo methods. We compare the results obtained from using conjugate and reference priors. Model selection based on the prequential conditional predictive ordinates is developed.
متن کاملBayesian Treatments for Panel Data Stochastic Frontier Models with Time Varying Heterogeneity
This paper considers a linear panel data model with time varying heterogeneity. Bayesian inference techniques organized around Markov chain Monte Carlo (MCMC) are applied to implement new estimators that combine smoothness priors on unobserved heterogeneity and priors on the factor structure of unobserved effects. The latter have been addressed in a non-Bayesian framework by Bai (2009) and Knei...
متن کاملDensity Modeling and Clustering Using Dirichlet Diffusion Trees
I introduce a family of prior distributions over multivariate distributions, based on the use of a “Dirichlet diffusion tree” to generate exchangeable data sets. These priors can be viewed as generalizations of Dirichlet processes and of Dirichlet process mixtures, but unlike simple mixtures, they can capture the hierarchical structure present in many distributions, by means of the latent diffu...
متن کاملDefault Analysis of Mixture Models using Expected
Consider observations Y , distributed according to a mixture of densities Y P k j=1 w j f(j j); where 0 w j 1, P w j = 1, k and j correspond to unknown parameters of the mixture. In the a Bayesian framework, it is not possible to perform a default statistical analysis of the mixture using non-proper priors, N , for the component parameters, since the posterior distribution of these do not exist...
متن کامل