Research Article
A Semisupervised Learning Scheme with Self-Paced Learning for Classifying Breast Cancer Histopathological Images
Table 5
Precision (Prec.), recall (R), and F1-score (F1) values for binary classification.
| Mag. factor | % of pseudolabels | Prec. (%) | R (%) | F1 (%) |
| 40X | (top-5%) | 99.50 | 99.23 | 99.38 | (top-10%) | 99.89 | 99.79 | 99.81 | (top-20%) | 99.50 | 99.21 | 99.36 | All pseudolabels | 98.72 | 98.63 | 98.49 |
| 100X | (top-5%) | 99.73 | 99.58 | 99.69 | (top-10%) | 99.28 | 99.17 | 99.23 | (top-20%) | 98.62 | 98.24 | 98.71 | All pseudolabels | 99.12 | 99.06 | 99.19 |
| 200X | (top-5%) | 99.43 | 98.91 | 99.18 | (top-10%) | 99.84 | 99.80 | 99.49 | (top-20%) | 99.27 | 99.00 | 99.13 | All pseudolabels | 99.18 | 99.10 | 99.22 |
| 400X | (top-5%) | 99.40 | 99.17 | 99.20 | (top-10%) | 99.85 | 99.77 | 99.54 | (top-20%) | 99.25 | 99.18 | 99.21 | All pseudolabels | 99.20 | 99.00 | 99.14 |
|
|