The LASSO performed variable selection as expected, by preferentially selecting variables associated with outcome over noise variables, and by preferring the strongest individually representative feature from within each correlated block, identifying strongly associated features more frequently than weakly associated features (Supplementary Fig. S1).
The feature selection method chosen was univariate analysis selecting variables with (P < 0.2) to add to the final model, and used the Akaike Information Criterion (AIC) [37] to identify the best model.
Models were built by using purposeful selection, a 5-step method for selecting variables based on both biologic importance and statistical significance.
We chose not to use stepwise variable selection techniques, which have been criticized for selecting variables for inclusion in a model based on statistical rather than clinical significance (Williams et al, 2006).
On the other hand, covariate selection may seem arbitrary, which is in favour of selecting variables that have already been shown to have an impact on capture probability [ 18, 39, 21].
Modelling was done using a strategy of selecting variables based on epidemiological knowledge and no actual model selection based on significance was done.
More recently, fuzzy logic proved beneficial when selecting variables to serve as neural network inputs.
A multivariate logistic regression model was constructed by selecting variables found to be significantly associated with complications in a univariate analysis.
The problem of selecting variables or features in a regression model in the presence of both additive (vertical) and leverage outliers is addressed.
Models developed by selecting variables gave lower prediction errors and proved to be a good approach to limit the effect of changing raw material characteristics.
The experimental results showed that the proposed framework provided high prediction accuracy in selecting variables for accurate prediction of fish freshness and meat safety.
