On the marginal likelihood and cross-validation
From MaRDI portal
Publication:5113024
Abstract: In Bayesian statistics, the marginal likelihood, also known as the evidence, is used to evaluate model fit as it quantifies the joint probability of the data under the prior. In contrast, non-Bayesian models are typically compared using cross-validation on held-out data, either through -fold partitioning or leave--out subsampling. We show that the marginal likelihood is formally equivalent to exhaustive leave--out cross-validation averaged over all values of and all held-out test sets when using the log posterior predictive probability as the scoring rule. Moreover, the log posterior predictive is the only coherent scoring rule under data exchangeability. This offers new insight into the marginal likelihood and cross-validation and highlights the potential sensitivity of the marginal likelihood to the choice of the prior. We suggest an alternative approach using cumulative cross-validation following a preparatory training phase. Our work has connections to prequential analysis and intrinsic Bayes factors but is motivated through a different course.
Recommendations
- Bayesian Model Assessment and Comparison Using Cross-Validation Predictive Densities
- Information criteria and cross validation for Bayesian inference in regular and singular cases
- Cross-validation prior choice in Bayesian probit regression with many covariates
- A comparison of marginal likelihood computation methods
- Bayesian model selection and model averaging
Cited in
(20)- Combining data envelopment analysis and stochastic frontiers via a LASSO prior
- Asymptotic Properties of Adaptive Likelihood Weights by Cross-Validation
- Fast Cross-validation for Multi-penalty High-dimensional Ridge Regression
- Marginal Likelihood Estimation with the Cross-Entropy Method
- Maximum likelihood estimation and uncertainty quantification for Gaussian process approximation of deterministic functions
- Marginal Likelihood Computation for Model Selection and Hypothesis Testing: An Extensive Review
- Bayesian artificial neural networks for frontier efficiency analysis
- Asymptotic Bounds for Smoothness Parameter Estimates in Gaussian Process Interpolation
- Priors in Bayesian Deep Learning: A Review
- Assessment of generalised Bayesian structural equation models for continuous and binary data
- Bayesian learning in performance. Is there any?
- Adaptation of the tuning parameter in general Bayesian inference with robust divergence
- The no-free-lunch theorems of supervised learning
- A general approximation to nested Bayes factors with informed priors
- Evidential Calibration of Confidence Intervals
- Lower-dimensional posterior density and cluster summaries for overparameterized Bayesian models
- Unbiased and consistent nested sampling via sequential Monte Carlo
- Comparing scale parameter estimators for Gaussian process interpolation with the Brownian motion prior: leave-one-out cross validation and maximum likelihood
- Asymptotics for a class of parametric martingale posteriors
- Natural variational annealing for multimodal optimization
This page was built for publication: On the marginal likelihood and cross-validation
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5113024)
