研究文章
一个有效的预测糖尿病疾病基于深层神经网络系统
|
| 方法 |
评价指标 |
| 用默认值 |
用最优值 |
| Acc |
规范 |
Sens |
精准医疗 |
F1-score |
Acc |
规范 |
Sens |
精准医疗 |
F1-score |
|
| LR |
78.25 |
92.0 |
54.42 |
80.0 |
64.77 |
79.0 |
91.63 |
57.14 |
80.0 |
66.66 |
| 支持向量机 |
81.25 |
93.02 |
61.22 |
83.33 |
70.58 |
96.75 |
96.38 |
97.22 |
93.95 |
95.56 |
| XGBoost |
86.5 |
92.30 |
74.62 |
83.33 |
78.74 |
97.25 |
98.11 |
95.58 |
96.29 |
95.94 |
| DT |
98.00 |
96.77 |
100.0 |
94.93 |
97.40 |
98.75 |
98.81 |
98.59 |
97.90 |
98.24 |
| 射频 |
98.5 |
99.20 |
97.22 |
98.59 |
97.90 |
99.00 |
99.20 |
98.59 |
98.59 |
98.59 |
|
|