Corpus ID: 11292761

Calibration of conditional composite likelihood for Bayesian inference on Gibbs random fields

@inproceedings{Stoehr2015CalibrationOC,
  title={Calibration of conditional composite likelihood for Bayesian inference on Gibbs random fields},
  author={Julien Stoehr and Nial Friel},
  booktitle={AISTATS},
  year={2015}
}
Gibbs random elds play an important role in statistics, however, the resulting likelihood is typically unavailable due to an intractable normalizing constant. Composite likelihoods oer a principled means to construct useful approximations. This paper provides a mean to calibrate the posterior distribution resulting from using a composite likelihood and illustrate its performance in several examples. 
Bayesian Model Selection for Exponential Random Graph Models via Adjusted Pseudolikelihoods
TLDR
A method to adjust pseudolikelihoods to obtain a reasonable, yet tractable, approximation to the likelihood is specified, which allows implementation of widely used computational methods for evidence estimation and pursuit of Bayesian model selection of exponential random graph models for the analysis of social networks. Expand
Efficient Bayesian inference for exponential random graph models by correcting the pseudo-posterior distribution
TLDR
This work provides practical guidelines to correct a sample from such a pseudo-posterior distribution so that it is approximately distributed from the target posterior distribution and discusses the computational and statistical efficiency that result from this approach. Expand
Hidden Gibbs random fields model selection using Block Likelihood Information Criterion
TLDR
A new approximate version of the Bayesian Information Criterion is introduced that answers model choice questions such as the selection of the dependence structure or the number of latent states in Gibbs random fields, and the performances of BLIC are studied. Expand
Asymptotic normality, concentration, and coverage of generalized posteriors.
Generalized likelihoods are commonly used to obtain consistent estimators with attractive computational and robustness properties. Formally, any generalized likelihood can be used to define aExpand
Light and Widely Applicable MCMC: Approximate Bayesian Inference for Large Datasets
Light and Widely Applicable (LWA-) MCMC is a novel approximation of the Metropolis‐ Hastings kernel targeting a posterior distribution defined on a large number of observations. Inspired byExpand
Noisy Hamiltonian Monte Carlo for Doubly Intractable Distributions
TLDR
The behavior of HMC is studied when these quantities are replaced by Monte Carlo estimates to correct for doubly intractable distributions. Expand
Multi-resolution Multitask Gaussian Processes
We consider evidence integration from potentially dependent observation processes under varying spatio-temporal sampling resolutions and noise levels. We develop a multi-resolution multi-task (MRGP)Expand
Multi-resolution Multi-task Gaussian Processes
TLDR
This work develops shallow Gaussian Process (GP) mixtures that approximate the difficult to estimate joint likelihood with a composite one and deep GP constructions that naturally handle biases in the mean and generalize and outperform state of the art GP compositions and offer information-theoretic corrections and efficient variational approximations. Expand
Bayesian Few-Shot Classification with One-vs-Each Pólya-Gamma Augmented Gaussian Processes
TLDR
This work proposes a Gaussian process classifier based on a novel combination of Polya-gamma augmentation and the one-vs-each softmax approximation that allows us to efficiently marginalize over functions rather than model parameters. Expand

References

SHOWING 1-10 OF 19 REFERENCES
Bayesian inference for Gibbs random fields using composite likelihoods
  • N. Friel
  • Computer Science, Mathematics
  • Proceedings Title: Proceedings of the 2012 Winter Simulation Conference (WSC)
  • 2012
TLDR
The contribution of this paper is to examine the performance of a collection of composite likelihood approximations in the context of Bayesian inference. Expand
Bayesian Inference from Composite Likelihoods, with an Application to Spatial Extremes
Composite likelihoods are increasingly used in applications where the full likelihood is analytically unknown or computationally prohibitive. Although the maximum composite likelihood estimator hasExpand
Bayesian composite marginal likelihoods
This paper proposes and discusses the use of composite marginal like- lihoods for Bayesian inference. This approach allows one to deal with complex statistical models in the Bayesian framework, whenExpand
Computational Bayesian Analysis of Hidden Markov Models
Abstract Versions of the Gibbs Sampler are derived for the analysis of data from hidden Markov chains and hidden Markov random fields. The principal new development is to use the pseudolikelihoodExpand
Bayesian Inference in Hidden Markov Random Fields for Binary Data Defined on Large Lattices
Hidden Markov random fields represent a complex hierarchical model, where the hidden latent process is an undirected graphical structure. Performing inference for such models is difficult primarilyExpand
Learning with Blocks: Composite Likelihood and Contrastive Divergence
TLDR
This paper shows that composite likelihoods can be stochastically optimized by performing a variant of contrastive divergence with random-scan blocked Gibbs sampling, and demonstrates that using higher-order blocks improves both the accuracy of parameter estimates and the rate of convergence. Expand
AN OVERVIEW OF COMPOSITE LIKELIHOOD METHODS
A survey of recent developments in the theory and application of com- posite likelihood is provided, building on the review paper of Varin(2008). A range of application areas, includingExpand
Recursive computing and simulation-free inference for general factorizable models
SUMMARY We illustrate how the recursive algorithm of Reeves & Pettitt (2004) for general factorizable models can be extended to allow exact sampling, maximization of distributions and computation ofExpand
Efficient recursions for general factorisable models
Let n S-valued categorical variables be jointly distributed according to a distribution known only up to an unknown normalising constant. For an unnormalised joint likelihood expressible as a productExpand
Statistical Analysis of Non‐Lattice Data
In rather formal terms, the situation with which this paper is concerned may be described as follows. We are given a fixed system of n sites, labelled by the first n positive integers, and anExpand
...
1
2
...