Table 4 The ensemble models accuracies with a confidence interval of 95% obtained after ROC analysis using weighted voting
No. of feat | Class probabilities for all features using the mean ROC cutoff values (M1) (%) | Class probabilities using the corresponding ROC cutoff for each feature (M2) (%) | Crisp labels for all features using the mean ROC cutoff value (M3) (%) |
|---|---|---|---|
1 | 75.49 (66.32,83.0) | 75.49 (66.32,83.0) | 73.53 (64.23,81.0) |
2 | 76.47 (67.37,84.0) | 72.55 (63.19,80.0) | 73.53 (64.23,81.0) |
3 | 78.43 (69.5,85.0) | 73.53 (64.23,81.0) | 79.41 (70.57,86.0) |
4 | 77.45 (68.43,84.0) | 71.57 (62.16,79.0) | 74.51 (65.27,82.0) |
5 | 75.49 (66.32,83.0) | 68.63 (59.09,77.0) | 81.37 (72.73,88.0) |
6 | 73.53 (64.23,81.0) | 69.61 (60.1,78.0) | 79.41 (70.57,86.0) |
7 | 75.49 (66.32,83.0) | 71.57 (62.16,79.0) | 80.39 (71.65,87.0) |
8 | 73.53 (64.23,81.0) | 70.59 (61.13,79.0) | 80.39 (71.65,87.0) |
9 | 75.49 (66.32,83.0) | 72.55 (63.19,80.0) | 80.39 (71.65,87.0) |
10 | 76.47 (67.37,84.0) | 69.61 (60.1,78.0) | 80.39 (71.65,87.0) |
11 | 75.49 (66.32,83.0) | 69.61 (60.1,78.0) | 80.39 (71.65,87.0) |
12 | 77.45 (68.43,84.0) | 70.59 (61.13,79.0) | 79.41 (70.57,86.0) |
13 | 77.45 (68.43,84.0) | 71.57 (62.16,79.0) | 80.39 (71.65,87.0) |