Therefore, the final model selection will be determined by the F1-score. F1-scores are provided on the Classification Reports 4 and Appendix 2, Table 8, which indicate that the model with default hyperparameters has a 79.64% F1-score, which is 0.36% higher than the model with tuned hyperparameters 79.28%. Therefore, we conclude that tuning with hyperparameters did not help the enhancement of the F1-scores of the model and, hence, with choose GBC with an original dataset as our final model.