Table 11.
Experimental results on Cleveland dataset with 13 features
| Mean ± SD | RF | LR | SVM | ELM | KNN | Proposed ensemble |
|---|---|---|---|---|---|---|
| E (%) | 76.01 ± 5.39 | 77.29 ± 5.52 | 75.74 ± 6.15 | 68.29 ± 8.95 | 58.43 ± 4.32 | 82.07 ± 6.00 |
| Precision (%) | 74.23 ± 6.41 | 76.84 ± 5.14 | 75.16 ± 7.47 | 65.54 ± 11.57 | 50.26 ± 6.74 | 83.79 ± 7.59 |
| Recall (%) | 68.08 ± 7.92 | 69.40 ± 13.02 | 69.41 ± 12.68 | 56.75 ± 14.76 | 45.20 ± 7.59 | 75.88 ± 11.08 |
| G-mean | 71.05 ± 6.75 | 73.59 ± 6.58 | 71.61 ± 7.07 | 61.45 ± 12.32 | 49.71 ± 6.20 | 79.76 ± 7.76 |
| MC (%) | 87.19 ± 21.18 | 81.61 ± 27.83 | 82.08 ± 29.68 | 114.60 ± 37.19 | 152.39 ± 19.74 | 62.96 ± 26.52 |
| Specificity (%) | 74.50 ± 9.02 | 79.31 ± 9.11 | 74.80 ± 8.20 | 67.32 ± 11.32 | 49.20 ± 11.80 | 84.16 ± 6.70 |
| AUC (%) | 70.22 ± 7.74 | 72.18 ± 5.69 | 71.18 ± 7.73 | 66.75 ± 11.40 | 45.32 ± 8.33 | 79.53 ± 8.24 |
The average SD on 10-folds CV. The best result is bolded