Development of a Hemodialysis Data Collection and Clinical Information System and Establishment of an Intradialytic Blood Pressure/Pulse Rate Predictive Model
Table 4.
Summary of best performances for BP/PRprediction taskusing different inputfeatures inANN models (based on MAE).
|
| Features |
Training loss |
Validation loss |
Test loss |
|
|
| Only four artery-related features |
10.12 |
11.02 |
11.23 |
|
| Three specific features, excluding any artery-related features |
12.98 |
14.75 |
14.43 |
|
| Eight features (including HPCV) |
13.51 |
12.45 |
12.41 |
|
| Seven features (w/o HPCV) |
10.32 |
11.60 |
11.71 |
|
|
|
|
