Table 10 Wilcoxon signed-rank test: proposed ensemble versus individual classifiers on Cleveland dataset
From: A hybrid cost-sensitive ensemble for heart disease prediction
| Â | RF | LR | SVM | ELM | KNN |
|---|---|---|---|---|---|
E | |||||
Z-value\(^{1}\) | − 2.668 | − 2.312 | − 2.655 | − 2.810 | − 2.803 |
p value\(^{2}\) | \(0.004^{**}\) | \(0.021^{*}\) | \(0.008^{**}\) | \(0.002^{**}\) | \(0.002^{**}\) |
Precision | |||||
Z-value | − 2.533 | − 2.318 | − 2.671 | − 2.814 | − 2.809 |
p value | \(0.011^{*}\) | \(0.016^{*}\) | \(0.006^{**}\) | \(0.002^{**}\) | \(0.002^{**}\) |
Recall | |||||
Z-value | − 2.668 | − 2.173 | − 2.524 | − 2.668 | − 2.668 |
p value | \(0.004^{**}\) | \(0.031^{*}\) | \(0.008^{**}\) | \(0.004^{**}\) | \(0.004^{**}\) |
G-mean | |||||
Z-value | − 2.666 | − 2.310 | − 2.703 | − 2.803 | − 2.803 |
p value | \(0.004^{**}\) | \(0.020^{*}\) | \(0.004^{**}\) | \(0.002^{**}\) | \(0.002^{**}\) |
MC | |||||
Z-value | − 2.668 | − 2.312 | − 2.655 | − 2.810 | − 2.803 |
p value | \(0.004^{**}\) | \(0.020^{*}\) | \(0.006^{**}\) | \(0.002^{**}\) | \(0.002^{**}\) |
Specificity | |||||
Z-value | − 1.892 | − 1.696 | − 2.316 | − 2.521 | − 2.553 |
p value | 0.094 | 0.101 | \(0.023^{*}\) | \(0.008^{**}\) | \(0.008^{**}\) |
AUC | |||||
Z-value | − 2.521 | − 2.310 | − 2.666 | − 2.803 | − 2.803 |
p value | \(0.008^{**}\) | \(0.020^{*}\) | \(0.004^{**}\) | \(0.002^{**}\) | \(0.002^{**}\) |