研究文章
乳腺癌分类使用极端的基于机器学习从乳房x光成像DenseNet121模型
|
| 模型 |
不。的迭代 |
特异性 |
灵敏度 |
训练精度 |
测试精度 |
计算时间在几分钟内 |
|
| VGG19 |
One hundred. |
97.81 |
98.05 |
94.63 |
91.64 |
214.3589 |
| MobileNet |
One hundred. |
98.61 |
98.47 |
94.97 |
92.75 |
202.2457 |
| Xception |
One hundred. |
98.93 |
98.87 |
95.65 |
94.24 |
198.3321 |
| ResNet50V2 |
One hundred. |
98.64 |
98.44 |
95.95 |
94.67 |
197.3123 |
| InceptionV3 |
One hundred. |
One hundred. |
99.57 |
96.29 |
95.64 |
187.3125 |
| InceptionResNetV2 |
One hundred. |
98.67 |
One hundred. |
96.69 |
96.51 |
171.2487 |
| DenseNet201 |
One hundred. |
98.67 |
99.35 |
97.29 |
97.5 |
168.4589 |
| DenseNet121 |
One hundred. |
99.35 |
One hundred. |
98.41 |
97.97 |
164.5278 |
| DenseNet121 +榆树 |
One hundred. |
99.45 |
One hundred. |
99.34 |
98.53 |
163.8975 |
|
|
