Special: Model Selection
Index: The Book of Statistical Proofs ▷ Model Selection
Proofs and Definitions on Model Selection Criteria in the StatProofBook, as of 2023-12-23.
| Criterion | Definition | Derivation | Relationships | Other |
|---|---|---|---|---|
| Goodness-of- fit measures |
||||
| Residual variance | resvar | resvar-bias resvar-biasp resvar-unb |
||
| R-squared | rsq | rsq-der | rsq-mll | |
| Signal-to-noise ratio | snr | snr-rsq | ||
| Classical infor- mation criteria |
||||
| Akaike information criterion | aic aicc |
aicc-aic aicc-mll |
||
| Bayesian information criterion | bic | bic-der | ||
| Deviance information criterion | dev dic |
|||
| Bayesian model selection |
||||
| Model evidence | me lme uplme cvlme eblme vblme |
me-der lme-der |
lme-pnp | lme-anc lme-mean |
| Family evidence | fe lfe |
fe-der lfe-der |
lfe-lme | lfe-approx |
| Bayes factor | bf lbf |
lbf-der | lbf-lme | bf-trans bf-sddr bf-ep |
| Posterior model probability | pmp | pmp-der | pmp-bf pmp-lbf pmp-lme |
|
| Bayesian model averaging | bma | bma-der | bma-lme |