Estimating Bayes Factors via Posterior Simulation with the Laplace—Metropolis Estimator
利用后验模拟与Laplace–Metropolis估计量估计贝叶斯因子
The key quantity needed for Bayesian hypothesis testing and model selection is the integrated, or marginal, likelihood of a model. We describe a way to use posterior simulation output to estimate integrated likelihoods. We describe the basic Laplace—Metropolis estimator for models without random effects. For models with random effects, we introduce the compound Laplace-Metropolis estimator. We apply this estimator to data from the World Fertility Survey and show it to give…