Table 5 The accuracy of the proposed cost-sensitive method compared to other methods
Dataset | Classification Accuracy (% | ||||||
|---|---|---|---|---|---|---|---|
Baseline SVM | S-MVML-LA | GMEB | WJMI | IGIS | IWFS | Proposed | |
Iris | 77.33 | 83.11 (4) | 96.44 (2) | NA | 88 (3) | NA | 97.33 (1) |
Cancer | 95.13 | 96.71 (3) | 95.42 (5) | 100 (1) | 94.56 (6) | 100 (1) | 96.44 (4) |
Ionosphere | 64.67 | 76.92 (4) | 85.76 (2) | NA | 76.91 (3) | NA | 86.30 (1) |
Wine | 59.91 | 88.33 (2) | 97.37 (4) | NA | 93.31 (3) | NA | 98.86 (1) |
Pid | 64.67 | 64.97 (4) | 76.82 (2) | NA | 75.53 (3) | NA | 77.07 (1) |
Wdbc | 85.93 | 89.80 (4) | 94.89 (3) | NA | 95.78 (2) | NA | 97.01 (1) |
Musk | 83.61 | 85.50 (3) | 80.87 (6) | 84.99 (4) | 82.56 (5) | 85.79 (2) | 87.81 (1) |
Dermatology-6 | 94.41 | 96.53 (3) | 99.44 (2) | 88.87 (5) | 91.57 (4) | 87.28 (6) | 99.45 (1) |
FuelCons | 86.45 | 89.54 (3) | 89.73 (2) | 83.06 (4) | 78.12 (6) | 82.59 (5) | 90.73 (1) |
Movement_libras | 91.30 | 91.90 (3) | 97.13 (2) | 68.31 (6) | 80.62 (5) | 81.79 (4) | 97.18 (1) |
Sonar | 81.21 | 79.80 (4) | 86.09 (1) | 76.8 (5) | 75.41 (6) | 81.6 (3) | 86.09 (1) |
SPECTF | 79.42 | 79.35 (3) | 79.35 (3) | 78.93 (5) | 79.40 (2) | 80.4 (1) | 71.51 (6) |
Mnist | 98.51 | 95.97 (3) | 98.36 (2) | 75.17 (5) | NA | 79.35 (4) | 98.58 (1) |
Colon | 83.84 | 83.97 (3) | 69.48 (5) | 97.25 (2) | NA | 98.26 (1) | 83.97 (3) |
Caltch101 | 99.26 | 99.19 (1) | 99.14 (3) | 37.91 (4) | NA | 32.86 (5) | 99.19 (1) |
DLBCL77 | 88.23 | 75.33 (5) | 89.41 (4) | 100 (1) | NA | 100 (1) | 97.41 (3) |
Kddcup-rootkit-imap-vs-back | 99.01 | 98.88 (3) | 100 (1) | NA | NA | NA | 100 (1) |
Kddcup-buffer-overflow-vs-back | 98.65 | 97.93 (3) | 99.81 (2) | NA | NA | NA | 100 (1) |
Kddcup-guess-passwd-vs-satan | 99.87 | 98.78 (3) | 99.95 (1) | NA | NA | NA | 99.95 (1) |
Kddcup-land-vs-satan | 99.19 | 99.31 (3) | 100 (1) | NA | NA | NA | 100 (1) |
Cleveland | 81.64 | 81.65 (3) | 81.91 (2) | NA | NA | NA | 82.96 (1) |
Average rank | Â | 3.23 | 2.73 | 3.90 | 4.36 | 3.09 | 1.71 |