Research Article
Feature Selection and Parameter Optimization of Support Vector Machines Based on Modified Artificial Fish Swarm Algorithms
Table 6
Result of Experiment 2.
|
Dataset | Number of original features | Feature selection by AFSA | Feature selection by MAFSA | Number of selected features | Average accuracy rate (%) | Number of selected features | Average accuracy rate (%) |
| Bupa | 6 | 4.1 | 84.94 | 3.9* | 85.22** | Pima | 8 | 4.8 | 83.72 | 4.2* | 83.85** | Glass | 9 | 4.9* | 89.69 | 5.1 | 91.58** | Vowel | 10 | 7.8 | 100** | 7* | 100** | Heart | 13 | 6.1* | 97.03 | 6.2 | 97.77** | Australian | 14 | 7.1 | 93.62** | 5.7* | 93.33 | Vehicle | 18 | 11.2 | 91.01 | 11* | 92.08** | Robot | 24 | 7.6* | 96.57 | 8.3 | 97.25** | German | 24 | 14.2 | 83.7 | 13.7* | 84.6** | Sonar | 60 | 27.2* | 99.04 | 29.1 | 100** |
|
|
It indicates the fewest number of selected features. It indicates the highest classification accuracy.
|