Metraipolis › archive › dc_e22d0049e676

Widely applicable information criterion — Wikipedia (en)

Source
Wikipedia (en)
Original
https://en.wikipedia.org/wiki/Widely_applicable_information_criterion
License (as archived)
CC-BY-SA-4.0
Archived
2026-10-07T12:57:31.785Z
SHA-256 of the archived text
f0fe399fbaec19e95b186a304c5e7b898342063d4c63050a6ed3afdc97165e3a
Stable URL of this page
https://metraipolis.com/documento/dc_e22d0049e676

How to cite

Widely applicable information criterion — Wikipedia (en) — Wikipedia (en), https://en.wikipedia.org/wiki/Widely_applicable_information_criterion (CC-BY-SA-4.0); archived copy: https://metraipolis.com/documento/dc_e22d0049e676 (Metraipolis archive dc_e22d0049e676, sha256 f0fe399fbaec19e95b186a304c5e7b898342063d4c63050a6ed3afdc97165e3a)
BibTeX
@misc{dce22d0049e676,
  title = {Widely applicable information criterion — Wikipedia (en)},
  howpublished = {Wikipedia (en)},
  url = {https://en.wikipedia.org/wiki/Widely_applicable_information_criterion},
  note = {License CC-BY-SA-4.0; archived copy https://metraipolis.com/documento/dc_e22d0049e676; sha256 f0fe399fbaec19e95b186a304c5e7b898342063d4c63050a6ed3afdc97165e3a},
  year = {2026}
}

Archived text

Reproduced under CC-BY-SA-4.0. Attribution: Wikipedia (en), https://en.wikipedia.org/wiki/Widely_applicable_information_criterion. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

In statistical science, the widely applicable information criterion (WAIC), also known as the Watanabe–Akaike information criterion, is the generalized version of the Akaike information criterion (AIC) onto singular statistical models. It is used as measure of how well the model will predict data it wasn't trained on. It is asymptotically equivalent to cross-validation loss. Lower values of WAIC correspond to better performance. If we take log pointwise predictive density: lppd ( y , Θ ) = ∑ i log 1 S ∑ s p ( y i ∣ Θ s ) {\displaystyle {\text{lppd}}(y,\Theta )=\sum _{i}\log {\frac {1}{S}}\sum _{s}p(y_{i}\mid \Theta _{s})} Then: WAIC ( y , Θ ) = − 2 ( lppd − ∑ i Var θ log p ( y i ∣ θ ) ⏟ penalty term ) {\displaystyle {\text{WAIC}}(y,\Theta )=-2\left({\text{lppd}}-\underbrace {\sum _{i}\operatorname {Var} _{\theta }\log p(y_{i}\mid \theta )} _{\text{penalty term}}\right)} Where y {\displaystyle y} is predicted output in training data. Θ is models posterior distribution, s {\displaystyle s} are samples from posterior, and i iterates over training data. In other words, in Bayesian statistics the posterior is represented by list of samples from it. WAIC penalty is then the variance of predictions among these samples, calculated and added for each datapoint from dataset. The penalty term is often referred to as the "effective number of parameters". This terminology stems from historical conventions, as a similar term is used in the Akaike Information Criterion. Watanabe recommends in practice calculating both WAIC and PSIS – Pareto Smoothed Importance Sampling. Both are approximations of leave-one-out cross-validation. If they disagree then at least one of them is not reliable. Similarly PSIS can sometimes detect if its estimate is not reliable (if k ^ {\displaystyle {\hat {k}}} > 0.7). Some textbooks of Bayesian statistics recommend WAIC over other information criteria, especially for multilevel and mixture models. Widely applicable Bayesian information criterion (WBIC) is the generalized version of Bayesian information criterion (BIC) onto singular statistical models. WBIC is the average log likelihood function over the posterior distribution with the inverse temperature > 1/log n where n is the sample size. Both WAIC and WBIC can be numerically calculated without any information about a true distribution.

Tema degli esploratori: statistica e scienza dei dati (deposito sistematico; categoria «Bayesian statistics»).
Fonte: Wikipedia in inglese, voce «Widely applicable information criterion» — https://en.wikipedia.org/wiki/Widely_applicable_information_criterion
Wikidata Q30591892 (dati Wikidata in CC0 1.0) — https://www.wikidata.org/wiki/Q30591892
Autori: i contributori di Wikipedia (cronologia: https://en.wikipedia.org/wiki/Widely_applicable_information_criterion?action=history).
Licenza: Creative Commons Attribuzione-Condividi allo stesso modo 4.0 (CC BY-SA 4.0) — https://creativecommons.org/licenses/by-sa/4.0/. Il riuso richiede attribuzione e stessa licenza.
Estratto dell'introduzione non modificato nel contenuto; revisione 1364229468, consultata il 2026-10-07T12:28:02.006Z tramite l'API ufficiale di Wikipedia. Impronta SHA-256 dell'estratto: f541d3db8df8be96ce47c1217f993562d88946c466cb3eab84ab33f4bbb7e0b5