Marginal Neural Ratio Estimation

(no analogy exists in crianca summary)

Technical definition

An inference technique used to estimate marginal likelihoods for model selection, combining neural networks with Markov chain Monte Carlo methods.

Browse the Nebula Glossary