Table 8 Performance comparisons between our hybrid model and the other schemes on BreaKHis
From: Breast cancer histopathology image classification through assembling multiple compact CNNs
Acc. | Str. | Magnification Factors | |||
|---|---|---|---|---|---|
|  |  | 40 × | 100 × | 200 × | 400 × |
PL | [9] | 83.8 ±4.1 | 82.1 ±4.9 | 85.1 ±3.1 | 82.3 ±3.8 |
|  | [17] | 83.0 ±3.0 | 83.2 ±3.5 | 84.6 ±2.7 | 82.1 ±4.4 |
|  | [11] | 90.0 ±6.7 | 88.4 ±4.8 | 84.6 ±4.2 | 86.1 ±6.2 |
|  | Our | 85.2 ±2.6 | 83.5 ±3.8 | 84.1 ±1.4 | 79.3 ±2.7 |
IL | [9] | 82.8 ±3.6 | 80.7 ±4.9 | 84.2 ±1.6 | 81.2 ±3.6 |
|  | [11] | 85.6 ±4.8 | 83.5 ±3.9 | 83.1 ±1.9 | 80.8 ±3.0 |
|  | Our | 85.7 ±1.9 | 84.2 ±3.2 | 84.9 ±2.2 | 80.1 ±4.4 |