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. |
97.73 |
97.98 |
97.76 |
97.88 |
211.2344 |
| MobileNet |
One hundred. |
98.53 |
98.41 |
98.23 |
98.42 |
198.1212 |
| Xception |
One hundred. |
98.65 |
98.82 |
98.85 |
98.73 |
194.2076 |
| ResNet50V2 |
One hundred. |
98.56 |
98.87 |
98.76 |
98.37 |
193.1878 |
| InceptionV3 |
One hundred. |
One hundred. |
99.52 |
98.63 |
98.52 |
183.1881 |
| InceptionResNetV2 |
One hundred. |
98.59 |
One hundred. |
98.58 |
98.47 |
167.1242 |
| DenseNet201 |
One hundred. |
98.84 |
99.28 |
98.98 |
98.85 |
164.3344 |
| DenseNet121 |
One hundred. |
99.18 |
One hundred. |
99.27 |
98.84 |
160.4033 |
| DenseNet121 +榆树 |
One hundred. |
99.37 |
99.94 |
99.47 |
99.14 |
159.7731 |
|
|
