Table 2 Results of hyperparameter Optimization. The hyperparameters are sorted from lowest to highest based on the resulting model’s accuracy in each step
Survival Random Forest | Value | Accuracy | Survival XGboost | Value | Accuracy |
|---|---|---|---|---|---|
No. of trees | 1000 | 62.5% | No. of consecutive trees | 195 | 58.8% |
No. of variables in each split | 5 | 63.7% | Maximum depth of tree | 15 | 60.6% |
Min No. of objects in each terminal node | 27 | 65% | Minimum child weight | 7 | 61.8% |
Gamma | 0.35 | 63.5% | |||
Eta | 0.1 | 64% |