Marginal Likelihood From the Metropolis–Hastings Output

By Siddhartha Chib, Ivan Jeliazkov. 2001

1,146 citations recorded by the source corpus.

Subfields: Artificial Intelligence, Statistics and Probability

View Paper: https://www.semanticscholar.org/paper/e33e3ce9541f261dd1d1c35190cf191347c1c15e

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Artificial Intelligence

In Statistics and Probability

Citations in the corpus

Variational learning for latent Gaussian model of discrete data (2012, methodology): “The Chib estimator A simple method for estimating marginal likelihood using Gibbs sampling or MH algorithm is proposed by Chib [Chib, 1995; Chib and Jeliazkov, 2001].”
Bayesian Analysis of DSGE Models (2006, positive): “, Christiano and Eichenbaum (1992), minimum distance estimation based on the discrepancy among VAR and DSGE model impulse response functions, e.”
Causal network inference using biochemical kinetics (2014, positive): “To estimate marginal likelihoods from sampler output, we exploit partial conjugacy and use the method of Chib and Jeliazkov (2001). As inference in CheMA 1.”
Probit and Logit Model Selection (2010, methodology): “We compute the marginal likelihood using the method given by Chib and Jeliazkov ( 2001 ).”
Marginal likelihood estimation via power posteriors (2008, background): “Within-model computations essentially aim to estimate the marginal likelihood (2) for each model k separately, and then if desired use this infor mation to form Bayes factors (Chib, 1995; Chib and Jeliazkov, 2001).”
Markov chain Monte Carlo methods for stochastic volatility models (2002, methodology): “To estimate (Â * |y; * ) we note from Chib and Jeliazkov (2001) that the ordinate can be expressed as (Â * |y; * ) = (Â; Â * |y * ; {s t })q(Â; Â * |y * ; {s t }) d(Â; {s t }; {q t }; { t }; ÿ; { t }|y; * ) (Â * ; Â|y * ; {s t }) dq(Â * ; Â|y * ; {s t }) d({s t }; {q t }; { t }; ÿ; { t }|y; * )…”
An introduction to sampling via measure transport (2016, background): “ormalizing constant is a particularly useful quantity in the context of Bayesian model selection [30]. Reliably retrieving this normalizing constant from MCMC samples requires additional eort (e.g., [31,20]). The numerical solution of (11) entails at least two dierent approximations. First, the inni”
Periodic Stochastic Volatility and Fat Tails (2005, methodology): “The marginal likelihood input to the computation of Bayes factors is constructed as in Chib (1995), and Chib and Jeliazkov (2001).”
Analysis of high dimensional multivariate stochastic volatility models (2006, methodology): “In previous work within the ARCH-GARCH tradition, multivariate models of volatility have been discussed by Bollerslev, Engle, and Wooldridge (1988), Diebold and Nerlove (1989), Engle, Ng, and Rothschild (1990), Lin (1992), King, Sentana, and Wadhwani (1994), Engle and Kroner (1995), Braun, Nelson, a”
PASTIS: Bayesian extrasolar planet validation. I. General framework, models, and performance (2014, methodology): “We have recently started testing additional methods (e.g. Chib & Jeliazkov 2001) to study this potential issue.”
Bayesian Inference of General Linear Restrictions on the Cointegration Space (2007, neutral): “The resulting estimator is the harmonic mean of the likelihoods of the posterior draws; Geweke (1999) describes a modied harmonic estimator along the lines initially suggested by Gelfand and Dey (1994). Yet another use of posterior samples for computing marginal likelihoods has recently been propos”
Estimating the Integrated Likelihood via Posterior Simulation Using the Harmonic Mean Identity (2006, methodology): “Chib and Jeliazkov (2001) extended this to the case of the Metropolis-Hastings algorithm, in which case it requires a different auxiliary simulation algorithm additional to the main MCMC algorithm.”
An Economic Evaluation of Empirical Exchange Rate Models (2009, positive): “For the 2 parameters in GARCH and SV, which are sampled in the MCMC chain by implementing a Metropolis-Hastings algorithm, the posterior density is computed as in Chib and Jeliazkov (2001).”
THE KEPLER-10 PLANETARY SYSTEM REVISITED BY HARPS-N: A HOT ROCKY WORLD AND A SOLID NEPTUNE-MASS PLANET (2014, methodology): “In order to assess which model is best, we calculated the marginal likelihood of each model according to the method of Chib & Jeliazkov (2001) (see the Appendix in R. D. Haywood et al., in preparation), which uses the posterior distributions of the MCMC chain.”