Practical Bayesian model evaluation using leave-one-out cross-validation and WAIC

By Aki Vehtari, Andrew Gelman, Jonah Gabry. 2015

5,067 citations recorded by the source corpus.

Leave-one-out cross-validation (LOO) and the widely applicable information criterion (WAIC) are methods for estimating pointwise out-of-sample prediction accuracy from a fitted Bayesian model using the log-likelihood evaluated at the posterior simulations of the parameter values. LOO and WAIC have various advantages over simpler estimates of predictive error such as AIC and DIC but are less used in practice because they involve additional computational steps. Here we lay out fast and stable computations for LOO and WAIC that can be performed using existing simulation draws. We introduce an efficient computation of LOO using Pareto-smoothed importance sampling (PSIS), a new procedure for regularizing importance weights. Although WAIC is asymptotically equal to LOO, we demonstrate that PSIS-LOO is more robust in the finite case with weak priors or influential observations. As a byproduct of our calculations, we also obtain approximate standard errors for estimated predictive errors and for comparison of predictive errors between two models. We implement the computations in an R package called loo and demonstrate using models fit with the Bayesian inference package Stan.

Subfields: Artificial Intelligence, Control and Systems Engineering, Statistics and Probability, Probabilistic & Bayesian Methods

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

  1. Aki Vehtari
  2. Andrew Gelman
  3. Paul-Christian Burkner
  4. Yuling Yao
  5. Jonah Gabry
  6. Daniel P. Simpson
  7. Yong Luo
  8. Khaleel Al-Harbi

In Control and Systems Engineering

  1. Paul-Christian Burkner
  2. Aki Vehtari
  3. Juho Piironen
  4. M. Korevaar
  5. Yong Luo
  6. Khaleel Al-Harbi
  7. Marc K. Francke
  8. Yuling Yao

In Statistics and Probability

  1. Aki Vehtari
  2. Paul-Christian Burkner
  3. Andrew Gelman
  4. Yuling Yao
  5. Daniel P. Simpson
  6. Jonah Gabry
  7. Yong Luo
  8. Khaleel Al-Harbi

In Probabilistic & Bayesian Methods

  1. Paul-Christian Burkner
  2. Aki Vehtari
  3. Yong Luo
  4. Khaleel Al-Harbi
  5. Charles C. Margossian
  6. Yuling Yao
  7. Daniel P. Simpson
  8. Juho Piironen

Citations in the corpus