Table 6.

Comparison of global performance measures for the majority voting ensemble approach and the best individual classifiers (non-ordinal and ordinal).

Performance Measure	Majority Voting Model Based on Ordinal and Non-Ordinal Classifiers	Best Non-Ordinal Classifier: Logistic Regression	Best ordinal classifier: ordinal AdaBoost based on OBE$\mathbf{(}\mathit{v}\mathbf{(}{\mathit{c}}^{\mathit{m}\mathit{a}\mathit{x}}\mathbf{)}\mathbf{)}$
F-score	0.501	0.453	0.475
Accuracy	0.558	0.505	0.526
AUC	0.596	0.560	0.578
MSE	1.074	1.189	1.137
${\tau }_b$	0.200	0.060	0.016