The R package tree was used to train classifiers grown using recursive partitioning with a deviance criterion to select features at each node [47].
Regression trees are constructed using the "deviance" criterion.
The complement deviance criterion seeks a model that would perform well in out-of-sample prediction.
It plots values of the selection parameter λ as estimated by LLARRMA and SS in a representative set of 50 of the 1,000 simulations (using the complement deviance criterion for LLARRMA, which seeks to maximize out-of-sample predictions).
We will fit the model with an independent conditional covariance structure and an autoregressive structure and choose the best data-supported model based on the deviance information criteria or a related criterion.
Bayesian p-value, deviance information criteria (DIC) and Akaike's information criterion (AIC) were used to compare the models [14].
Following the criterion of improvement in the model goodness of fit and parsimony (Deviance Information Criteria) the relevant levels will be chosen, contextual covariates will be incorporated and the interaction between levels will be analysed.
In a Bayesian framework the Bayesian information criterion (BIC) or deviance information criterion (DIC) can be used [ 15, 16].
The models were compared using the Deviance Information Criteria (DIC) [20].
DIC = Deviance Information Criteria.
