Probability & Statistics Codexery

Prior probability

Last updated

Priors can be constructed from past experiments, expert opinion, or chosen for their mathematical properties.[1][2]

Quick Facts

Definition
Assumed probability distribution before evidence

Facts from the source article.

Key facts and history

Prior distributions may depend on their own parameters, called hyperparameters, which can themselves have hyperprior distributions.[3] For example, a beta distribution used as a prior for a Bernoulli parameter has hyperparameters α and β.[3] An uninformative prior expresses vague information and is sometimes called a flat or diffuse prior.[4] Jaynes argued that the Haldane prior p⁻¹(1-p)⁻¹ represents complete uncertainty about a probability.[5]

The Haldane prior gives most weight to the extremes p=0 and p=1.[6] An improper prior is one where the sum or integral of prior values need not be finite.[7] Model parameters' prior distributions often depend on their own parameters, and these hyperparameters can themselves be expressed as hyperprior distributions, allowing for many conditional levels of distributions known as hierarchical priors.[3] If the parameter space has a natural group structure that leaves the Bayesian state of knowledge invariant, priors can be built proportional to the Haar measure.[5]

For a random variable's mean and variance, statisticians may adopt an improper prior such as p(m, v) ~ 1/v (for v > 0) as an uninformative prior.[8]

What the research sources say

The book "Economic Decision-Making: Games, Econometrics and Optimisation" was published by North-Holland in December 1990.[9] The book "Bayesian Hierarchical Models" was published by Taylor & Francis Group in 2021.[10] The book "An introduction to Bayesian inference in econometrics" was published by Wiley in 1971.[11]

References

  1. Wikipedia, "Prior probability" (CC BY-SA 4.0), which cites: Robert, "The Bayesian Choice", Springer, 1994, p. 89–136 (original not read by us) en.wikipedia.org
  2. Wikipedia, "Prior probability" (CC BY-SA 4.0), which cites: Chaloner, "Bayesian Biostatistics", Marcel Dekker, 1996, p. 141–156; Mikkola, "Prior Knowledge Elicitation: The Past, Present, and Future", Bayesian Analysis, 2024; Icazatti, "PreliZ: A tool-box for prior elicitation", Journal of Open Source Software, September 2023, p. 5499; Zellner, "An Introduction to Bayesian Inference in Econometrics", John Wiley & Sons, 1971, p. 41–53; Price, "Uninformative priors for Bayes' theorem", AIP Conf. Proc., 2001, p. 379–391; Piironen, "Sparsity information and regularization in the horseshoe and other en.wikipedia.org
  3. Wikipedia, "Prior probability" (CC BY-SA 4.0), which cites: Congdon, "Bayesian Hierarchical Models", CRC Press, 2020, p. 253–315 (we also read that source) en.wikipedia.org
  4. Wikipedia, "Prior probability" (CC BY-SA 4.0), which cites: Zellner, "An Introduction to Bayesian Inference in Econometrics", John Wiley & Sons, 1971, p. 41–53 (we also read that source) en.wikipedia.org
  5. Wikipedia, "Prior probability" (CC BY-SA 4.0), which cites: Jaynes, "Prior Probabilities", IEEE Transactions on Systems Science and Cybernetics, Sep 1968, p. 227–241 (original not read by us) en.wikipedia.org
  6. Wikipedia, "Prior probability" (CC BY-SA 4.0), which cites: "A note on inverse probability", Mathematical Proceedings of the Cambridge Philosophical Society, 1932, p. 55–61 (original not read by us) en.wikipedia.org
  7. Wikipedia, "Prior probability" (CC BY-SA 4.0), which cites: Dawid, "Marginalization Paradoxes in Bayesian and Structural Inference", Journal of the Royal Statistical Society, 1973, p. 189–233 (original not read by us) en.wikipedia.org
  8. Wikipedia, "Prior probability" (CC BY-SA 4.0), which cites: Christensen, "Bayesian Ideas and Data Analysis : An Introduction for Scientists and Statisticians", CRC Press, 2010, p. 69 (original not read by us) en.wikipedia.org
  9. Open Library record, ISBN 044488422X ("Economic Decision-Making: Games, Econometrics and Optimisation "). CC0 (metadata); book content not read. openlibrary.org
  10. Open Library record, ISBN 9781032177151 ("Bayesian Hierarchical Models"). CC0 (metadata); book content not read. openlibrary.org
  11. Open Library record, ISBN 0471981656 ("An introduction to Bayesian inference in econometrics."). CC0 (metadata); book content not read. openlibrary.org

Each numbered sentence above was checked against the source it cites before publication. Sentences citing Wikipedia adapt text from the English Wikipedia article Prior probability (contributors), used under CC BY-SA 4.0; this entry is available under the same licence. Other sources were used for facts only; no text or images are copied from them.

More in Probability & Statistics

Sources

Compiled from Wikipedia and the sources listed below. Text from Wikipedia is available under CC BY-SA 4.0; this entry is adapted from it.

Spotted an error? Know more?

Reader corrections go straight into our review queue. Suggest an edit · How this site is sourced

Comments

Loading…
Open in the interactive codex →