Table 7 Comparative analysis of five evolutionary algorithms and the proposed method.
From: Discriminative biomarker selection using hybrid multi-population evolutionary computation
Dataset | Measures | GA | PSO | DE | TLBO | GSA | Proposed |
|---|---|---|---|---|---|---|---|
Leukemia cancer | Acc ± Std | 86.25 ± 1.45 | 84.52 ± 1.25 | 87.23 ± 1.85 | 90.25 ± 2.01 | 92.65 ± 1.88 | 94.71 ± 0.57 |
#feat | 39 | 41 | 34 | 29 | 27 | 17 | |
ETime | 845.25 | 945.25 | 1120.52 | 1554.23 | 1104.27 | 711.35 | |
MLL | Acc ± Std | 80.25 ± 2.54 | 82.52 ± 2.87 | 88.30 ± 3.07 | 87.41 ± 2.98 | 88.94 ± 2.47 | 91.63 ± 2.06 |
#feat | 34 | 32 | 29 | 37 | 25 | 18 | |
ETime | 622.38 | 751.02 | 811.25 | 965.32 | 888.25 | 711.02 | |
Ovarian | Acc ± Std | 90.21 ± 2.14 | 91.25 ± 2.98 | 88.72 ± 2.45 | 89.65 ± 1.69 | 88.14 ± 1.35 | 93.66 ± 1.02 |
#feat | 25 | 24 | 21 | 29 | 21 | 14 | |
ETime | 1132.05 | 1223.52 | 1377.95 | 1425.08 | 1423.57 | 966.52 | |
CNS | Acc ± Std | 69.52 ± 1.25 | 74.85 ± 2.31 | 76.33 ± 3.01 | 75.98 ± 0.99 | 77.25 ± 1.24 | 81.65 ± 0.45 |
#feat | 30 | 34 | 28 | 27 | 25 | 18 | |
ETime | 877.98 | 811.63 | 799.25 | 1000.2 | 975.87 | 623.25 | |
SRBCT | Acc ± Std | 95.63 ± 1.32 | 94.85 ± 1.02 | 96.35 ± 0.96 | 93.02 ± 1.88 | 94.25 ± 0.91 | 99.52 ± 0.38 |
#feat | 19 | 21 | 22 | 17 | 27 | 11 | |
ETime | 889.25 | 966.37 | 1552.74 | 966.52 | 877.63 | 844.25 | |
Lung cancer | Acc ± Std | 94.25 ± 0.96 | 92.22 ± 1.36 | 91.52 ± 0.96 | 93.52 ± 0.68 | 91.55 ± 1.44 | 95.22 ± 0.31 |
#feat | 22 | 19 | 31 | 37 | 32 | 15 | |
ETime | 788.63 | 688.52 | 744.25 | 975.25 | 833.41 | 522.85 |
