Table 3 Average classification accuracy rate and as standard deviation (shown in parenthesis) over ten runs of the evolutionary-based feature selection methods using KNN, SVM, and AdaBoost classifier
From: A novel community detection based genetic algorithm for feature selection
Dataset | Method | Classifier | ||
|---|---|---|---|---|
KNN | SVM | AdaBoost | ||
SpamBase | PSO | 92.54 (2.83) | 92.35 (1.52) | 92.43 (2.25) |
ACO | 91.81 (1.82) | 89.51 (2.81) | 90.89 (2.21) | |
ABC | 90.35 (2.39) | 89.22 (1.93) | 91.30 (3.33) | |
CDGAFS | 93.99 (2.76) | 93.68 (1.73) | 93.27 (2.82) | |
Sonar | PSO | 88.18 (2.43) | 87.81 (2.29) | 86.93 (3.32) |
ACO | 88.06 (2.32) | 87.36 (3.32) | 85.82 (1.48) | |
ABC | 87.23 (1.13) | 87.17 (2.81) | 86.74 (1.78) | |
CDGAFS | 88.71 (3.76) | 88.34 (2.19) | 87.13 (2.71) | |
Arrhythmia | PSO | 86.15 (2.82) | 86.01 (2.61) | 85.91 (2.82) |
ACO | 84.13 (2.12) | 86.27 (2.62) | 85.72 (3.94) | |
ABC | 85.83 (2.73) | 85.71 (1.75) | 84.32 (1.39) | |
CDGAFS | 87.21 (2.37) | 87.38 (2.02) | 86.98 (2.59) | |
Madelon | PSO | 86.46 (3.14) | 86.65 (2.47) | 86.12 (1.81) |
ACO | 86.19 (2.20) | 85.91 (1.32) | 86.34 (2.11) | |
ABC | 87.55 (2.13) | 87.19 (1.81) | 86.12 (2.33) | |
CDGAFS | 87.88 (1.55) | 87.82 (1.64) | 86.79 (1.62) | |
Isolet | PSO | 85.63 (1.39) | 85.39 (1.62) | 85.42 (2.32) |
ACO | 85.26 (1.58) | 85.90 (1.81) | 85.41 (2.39) | |
ABC | 84.38 (2.81) | 84.95 (2.16) | 84.84 (1.48) | |
CDGAFS | 86.11 (2.44) | 86.01 (2.65) | 85.39 (2.62) | |
Colon | PSO | 96.41 (2.82) | 96.19 (2.16) | 96.32 (1.31) |
ACO | 94.43 (1.71) | 95.73 (1.19) | 95.32 (1.82) | |
ABC | 93.04 (2.56) | 92.61 (3.61) | 92.49 (3.45) | |
CDGAFS | 95.41 (2.15) | 95.82 (2.65) | 95.36 (2.38) | |