Easily Computed Marginal Likelihoods from Posterior Simulation Using the THAMES Estimator

2023/05/15 by Martin Metodiev, Marie Perrot-Dockès, Metodiev, Martin +7 · 3 citations
Decision Sciences · Mathematics · #Probability and Risk Models #Statistical Methods and Bayesian Inference #Statistical Methods and Inference

paper · pdf · doi:10.48550/arxiv.2305.08952

Abstract

We propose an easily computed estimator of marginal likelihoods from posterior simulation output, via reciprocal importance sampling, combining earlier proposals of DiCiccio et al (1997) and Robert and Wraith (2009). This involves only the unnormalized posterior densities from the sampled parameter values, and does not involve additional simulations beyond the main posterior simulation, or additional complicated calculations. It is unbiased for the reciprocal of the marginal likelihood, consistent, has finite variance, and is asymptotically normal. It involves one user-specified control parameter, and we derive an optimal way of specifying this. We illustrate it with several numerical examples.

Cited by

• Quantitative mobile gamma-ray spectrometry through Bayesian inference
• Observational constraints on New Tsallis holographic energy in Rastall theory
• Approximating evidence via bounded harmonic means

Related

• On the computation of the marginal likelihood
• Simulation-efficient marginal posterior estimation with swyft: stop wasting your precious time
• Marginal Likelihood Computation via Arrogance Sampling
• Computing the Bayesian Factor from a Markov chain Monte Carlo Simulation of the Posterior Distribution
• Sandwiching the marginal likelihood using bidirectional Monte Carlo