Research Article
A Study of Machine-Learning Classifiers for Hypertension Based on Radial Pulse Wave
Table 4
Selected (top 12) features in model.
| model | Feature variable | Feature importance |
| Adaboost | t | 0.126 | BMI | 0.109 | HR | 0.083 | h1/t1 | 0.066 | Ad | 0.057 | w2 | 0.051 | t2 | 0.049 | t3 | 0.049 | w1/t | 0.040 | h5 | 0.034 | h3/h1 | 0.034 | As | 0.034 |
| Gradient Boosting | H20 score | 0.128 | BMI | 0.075 | t | 0.065 | h1/t1 | 0.055 | t3 | 0.048 | h5 | 0.042 | t5 | 0.042 | Ad | 0.041 | t2 | 0.041 | t1 | 0.039 | h4 | 0.039 | w2/t | 0.039 |
|
|