Development of a Hemodialysis Data Collection and Clinical Information System and Establishment of an Intradialytic Blood Pressure/Pulse Rate Predictive Model
Table 5.
Summary of best performances for BP/PRprediction taskusing different inputfeatures in LSTM models (based on MAE).
|
| Features |
Training loss |
Validation loss |
Test loss |
|
| |
| All 15 features |
2.94 |
12.11 |
12.15 |
|
| Only 13 features |
3.29 |
12.06 |
11.28 |
|
| Only four artery-related features |
3.55 |
12.36 |
12.25 |
|
| Three specific features, excluding any artery-related features |
12.98 |
16.17 |
15.84 |
|
|
|
|
