Research Article
Correlation-Based Ensemble Feature Selection Using Bioinspired Algorithms and Classification Using Backpropagation Neural Network
Table 6
Feature importance of WDBC dataset.
| S. no. | Feature | Feature importance | Rank |
| 1 | ID | 0.852635 | 24 | 2 | Radius (mean) | 0.860782 | 22 | 3 | Radius (error) | 0.835712 | 27 | 4 | Radius (worst) | 0.926704 | 10 | 5 | Texture (mean) | 0.928031 | 8 | 6 | Texture (error) | 0.776179 | 29 | 7 | Texture (worst) | 0.909129 | 16 | 8 | Perimeter (mean) | 0.93506 | 2 | 9 | Perimeter (error) | 0.94209 | 1 | 10 | Perimeter (worst) | 0.735037 | 30 | 11 | Area (mean) | 0.836177 | 26 | 12 | Area (error) | 0.933734 | 5 | 13 | Area (worst) | 0.864297 | 20 | 14 | Smoothness (mean) | 0.931545 | 6 | 15 | Smoothness (error) | 0.925377 | 11 | 16 | Smoothness (worst) | 0.93505 | 3 | 17 | Compactness (mean) | 0.923189 | 12 | 18 | Compactness (error) | 0.928030 | 9 | 19 | Compactness (worst) | 0.858593 | 23 | 20 | Concavity (mean) | 0.818137 | 28 | 21 | Concavity (error) | 0.917486 | 14 | 22 | Concavity (worst) | 0.900307 | 17 | 23 | Concave (mean) | 0.863435 | 21 | 24 | Concave (error) | 0.898584 | 18 | 25 | Concave (worst) | 0.935045 | 4 | 26 | Symmetry (mean) | 0.719719 | 31 | 27 | Symmetry (error) | 0.918347 | 13 | 28 | Symmetry (worst) | 0.930219 | 7 | 29 | Fractal dimension (mean) | 0.914832 | 15 | 30 | Fractal dimension (error) | 0.845395 | 25 | 31 | Fractal dimension (worst) | 0.891554 | 19 |
|
|