Table 4 F1, precision, and recall on BreaKHis
From: Breast cancer histopathology image classification through assembling multiple compact CNNs
Cri. | Str. | Magnification Factors | |||
|---|---|---|---|---|---|
|  |  | 40 × | 100 × | 200 × | 400 × |
F1 | 1 | 86.7 ±2.2 | 86.6 ±0.6 | 86.4 ±2.1 | 83.5 ±2.9 |
|  | 2 | 89.4 ±0.8 | 88.6 ±1.7 | 89.0 ±1.7 | 86.3 ±3.2 |
|  | 3 | 89.8 ±1.0 | 88.8 ±1.5 | 89.4 ±1.3 | 86.8 ±2.8 |
Pr. | 1 | 85.4 ±1.7 | 82.5 ±0.9 | 85.2 ±2.1 | 79.2 ±3.7 |
|  | 2 | 84.7 ±1.7 | 83.7 ±5.1 | 87.5 ±2.3 | 81.2 ±5.7 |
|  | 3 | 85.3 ±1.8 | 83.9 ±4.7 | 88.0 ±1.5 | 81.6 ±5.1 |
Rec. | 1 | 88.3 ±4.0 | 91.3 ±2.1 | 87.7 ±4.8 | 88.5 ±3.3 |
|  | 2 | 94.8 ±1.2 | 94.5 ±3.0 | 90.6 ±2.0 | 92.5 ±3.2 |
|  | 3 | 94.9 ±0.8 | 94.6 ±2.9 | 90.8 ±1.9 | 93.1 ±3.0 |