For predictor pairs with Pearson r ≥0.9, we only retained one of the variables for the modeling [48] by selecting the variable with the strongest biological interpretability and the smallest correlation to the other predictor variables (Tables 1, 2).
Moreover, feature selection method can improve the predictive performance of models by selecting the predictor variables most related to target variable; in addition, feature selection can solve the problem when there are correlated variables in the data set which harm the performance of the models.
The development of the fuzzy expert system was carried out by selecting the input variables associated with the brine heater operation, building the fuzzy membership functions for input and output variables, and developing knowledge based rule structures.
Next, classification and regression trees (CART) [ 19] were composed that divided the sample in subgroups as homogeneous as possible with regard to the outcome by selecting the most predictive variables based on minimum prediction error.
To reduce the number of variables to be entered into one overall regression analysis to produce the final model predicting QoL-AD scores, the following method was adopted, which maximised the potential for a strong prediction by selecting the most predictive variables from each domain, and allowed identification of the domains which appeared most relevant to QoL-AD scores.
We therefore also performed a further selection of the 17 583 exon array probe sets, by selecting the 5000 most variable probe sets of these 17 583 (1.74% of the total number of 'core' probe sets, Figure 2B).
The CART algorithm grows a tree from the root by selecting the best predictor variable, which is the one with the lowest Gini index value, as an internal node.
Subsequently, a multi-objective optimization is carried out by employing a genetic algorithm that minimizes the compression energy and the required membrane area by selecting the optimal process layout, operating variables, and membrane materials.
