Table 7 Results summary comparing model-based (LASSO) and model-free (RF) prediction on testing data
From: Modeling and prediction of pressure injury in hospitalized patients using artificial intelligence
Metrics | Logistics regression | Random forest | ||||||
|---|---|---|---|---|---|---|---|---|
Design | Using the entire data | Without Braden metrics | Using the entire data | Without Braden metrics | ||||
Balance | Without balancing | With balancing | Without balancing | With balancing | Without balancing | With balancing | Without balancing | With balancing |
Accuracy | 0.9567 | 0.86 | 0.953 | 0.84 | 0.985 | 0.99 | 0.957 | 0.994 |
Kappa | 0.3871 | 0.69 | 0.302 | 0.642 | 0.813 | 0.979 | 0.2496 | 0.987 |
Sensitivity | 0.677 | 0.767 | 0.6129 | 0.72 | 0.72 | 1 | 0.163 | 1 |
Specificity | 0.963 | 0.912 | 0.959 | 0.9 | 0.998 | 0.9855 | 0.996 | 0.99 |
Pos pred val | 0.289 | 0.823 | 0.217 | 0.8 | 0.95 | 0.973 | 0.677 | 0.98 |
Neg pred val | 0.992 | 0.88 | 0.993 | 0.35 | 0.986 | 1 | 0.959 | 1 |
Balanced accuracy | 0.82 | 0.83 | 0.788 | 0.813 | 0.859 | 0.993 | 0.579 | 0.996 |
F1 score | 0.40 | 0.79 | 0.32 | 0.76 | 0.82 | 0.99 | 0.26 | 0.99 |