A relation between log-likelihood and cross-validation log-scores

It is shown that the log-likelihood of a hypothesis or model given some data is equal to an average of all leave-one-out cross-validation log-scores that can be calculated from all subsets of the data. This relation can be generalized to any k-fold cross-validation log-scores.

• Publication year

  2019

• Venue

  arXiv.org

• Publication date

  2019-08-18

• Fields of study

  Mathematics, Computer Science

• Identifiers
  DOI 10.31219/osf.io/k8mj3arXiv 1908.08741
• External record

  Open on Semantic Scholar

• Source metadata

  Semantic Scholar

• No claims are published for this paper.

• No concepts are published for this paper.

• Limitations of Bayesian Leave-One-Out Cross-Validation for Model Selection

  2018cited by this paper
• Bayesian model comparison: Log scores and DIC

  2014cited by this paper
• Understanding predictive information criteria for Bayesian models

  2013influential reference
• A survey of Bayesian predictive methods for model assessment, selection and comparison

  2012cited by this paper
• Bayesian Model Specification: Some problems related to model choice and calibration

  2011cited by this paper
• BAYESIAN THEORY

  2008cited by this paper
• Weight of Evidence : A Brief Survey

  2006cited by this paper
• Probability theory: the logic of science

  2005influential reference
• Bayesian Model Assessment and Comparison Using Cross-Validation Predictive Densities

  2002cited by this paper
• Bayes factors

  1995cited by this paper
• Bayesian Interpolation

  1992influential reference
• Bayesian Statistics 2.

  1987cited by this paper
• Good Thinking: The Foundations of Probability and Its Applications

  1983cited by this paper
• Some Logic and History of Hypothesis Testing

  1981cited by this paper
• A Predictive Approach to Model Selection

  1979cited by this paper
• An Asymptotic Equivalence of Choice of Model by Cross‐Validation and Akaike's Criterion

  1977influential reference
• Explicativity, corroboration, and the relative odds of hypotheses

  1975cited by this paper
• Information, weight of evidence, the singularity between probability measures and signal detection

  1974cited by this paper
• A Subjective Evaluation of Bode's Law and an ‘Objective’ Test for Approximate Numerical Rationality

  1969cited by this paper
• Probability and the Weighing of Evidence

  1950influential reference

• Don't guess what's true: choose what's optimal. A probability transducer for machine-learning classifiers

  2023cites this paper