New answers tagged

1 vote

Model selection: Is AIC enough or should one compute the p-value in model selection (and if yes to how to do it?)?

You've got the formula for the AIC wrong. The Akaike information criterion (AIC) is defined as: $$ \begin{aligned} \operatorname{AIC} = -2 \log\left\{\ell(\hat{\theta} | y)\right\} + 2k \end{...
user avatar
  • 2,722
1 vote

Model selection in simultaneous ARMA-GARCH modeling without AIC

Model selection should be guided by an explicit goal. Different goals may justify different models. Among information criteria, there are several alternatives AIC such as BIC, HQIC and more. They have ...
user avatar
2 votes

Model selection criteria that represent a compromise between AIC and BIC

There is the Hannan–Quinn information criterion (HQIC): $$ \text{HQ}=-2L_{\text{max}}+2k\ln(\ln(n)) $$ where $L_{\text{max}}$ is the maximized log-likelihood, $k$ is the number of parameters, and $n$ ...
user avatar
2 votes

AIC and BIC in GEE

The AIC does not exist for GEE models, because AIC is calculated from the likelihood and GEE models are not derived from a likelihood. Similarly the BIC does not exist for GEE models. Instead, the QIC ...
user avatar

Top 50 recent answers are included