Table 7 P-values from Friedman and Nemenyi tests across four models on three evaluation metrics.
Test Type | Model Comparison | C-index | ACC | AUC |
|---|---|---|---|---|
Friedman Test (4 models) | \(2.328 \times 10^{-6}\) | 0.021 | \(5.556 \times 10^{-6}\) | |
Nemenyi Test | Cox vs. RSF | 0.307 | 0.954 | 0.072 |
Cox vs. GBM | 0.002 | 0.160 | 0.029 | |
Cox vs. XGBoost | \(3.000 \times 10^{-6}\) | 0.954 | 0.986 | |
RSF vs. GBM | 0.226 | 0.046 | \(1.200 \times 10^{-6}\) | |
RSF vs. XGBoost | 0.006 | 1.000 | 0.029 | |
GBM vs. XGBoost | 0.508 | 0.046 | 0.072 | |
