Validation and performance evaluation
Validation of the algorithms were performed using five-fold
cross-validation method. Hyperparameters of the interested models were
tuned by random search optimization in the current study. We computed
the following performance metrics to measure the quality of a classifier
using five -fold cross-validation: Accuracy, sensitivity, specificity,
positive/negative predictive values, F-measure, and G-mean. All
performance metrics were calculated through DTROC software. The
supervised ML algorithms mentioned earlier were implemented using R
Studio of R programming language.