Table 4.
Estimated coefficient, standard error, Z-score and P-value from applying the Cox regression model to estimated predicted risk scores in the validation set obtained by OOR, LASSO, Ridge regression, Elastic net, Random forest, and generalized boosting models.
| Method | Coef | SE | Z-score | p-value |
|---|---|---|---|---|
| Object-Oriented Regression | 0.584 | 0.24 | 2.435 | 0.015 |
| LASSO | 0.712 | 0.716 | 0.995 | 0.320 |
| Ridge Regression | 0.000 | 0.001 | 0.218 | 0.827 |
| Elastic Net | 0.166 | 0.128 | 1.295 | 0.195 |
| Random Forest | 0.011 | 0.018 | 0.629 | 0.529 |
| Generalized Boosting | 0.676 | 0.334 | 2.026 | 0.043 |