Research Article
A Fast Hybrid Classification Algorithm with Feature Reduction for Medical Images
Table 3
Impact of classification accuracy using previous feature selection versus all features on different cervical cells.
| Cell category | All features | Previous 13 features | Previous 7 features | Precision | Recall | -score | Precision | Recall | -score | Precision | Recall | -score |
| Normal | 0.87 | 0.87 | 0.84 | 0.95 | 0.94 | 0.97 | 0.98 | 0.95 | 0.95 | Intermediate | 0.85 | 0.74 | 0.78 | 0.89 | 0.84 | 0.87 | 0.98 | 0.96 | 0.96 | Columnar | 0.77 | 0.77 | 0.75 | 0.95 | 0.95 | 0.98 | 0.99 | 0.94 | 0.99 | In situ | 0.74 | 0.87 | 0.78 | 0.94 | 0.97 | 0.94 | 0.98 | 0.90 | 0.97 | Preliminary dysplastic | 0.87 | 0.78 | 0.84 | 0.99 | 0.94 | 0.94 | 0.984 | 0.97 | 0.97 | Intermediate dysplastic | 0.79 | 0.85 | 0.77 | 0.87 | 0.97 | 0.94 | 0.986 | 0.97 | 0.97 | Final stage dysplastic | 0.79 | 0.75 | 0.70 | 0.87 | 0.95 | 0.94 | 0.988 | 0.94 | 0.95 |
|
|