Research Article
The Construction of Support Vector Machine Classifier Using the Firefly Algorithm
Table 2
The CCR of three different algorithms without feature selection (mean ± SD).
| Data set | Firefly-SVM | PSO-SVM | LIBSVM |
| SPECTF heart | 84.27 ± 0.95 | 80.19 ± 2.44 | 80.19 ± 2.73 | Breast Cancer Wisconsin (diagnosis) | 98.44 ± 0.39 | 97.54 ± 0.90 | 97.54 ± 1.54 | Statlog (heart) | 86.75 ± 2.19 | 86.75 ± 2.0 | 85.25 ± 2.99 | German credit data | 77.60 ± 0.89 | 75.33 ± 2.61 | 76.34 ± 1.24 | Sonar | 93.46 ± 3.31 | 92.69 ± 4.18 | 89.23 ± 4.63 | Pima-Indians-diabetes | 77.87 ± 1.33 | 76.34 ± 2.59 | 72.25 ± 3.22 | Australian credit approval | 88.63 ± 1.40 | 84.34 ± 3.77 | 83.76 ± 3.36 | Live disorders | 75.36 ± 2.99 | 73.22 ± 3.28 | 69.86 ± 2.67 | Ionosphere | 96.87 ± 1.91 | 95.21 ± 2.84 | 94.35 ± 2.25 | Breast Cancer Wisconsin (original) | 97.60 ± 0.37 | 96.62 ± 0.96 | 95.52 ± 0.92 |
|
|