Table 1. Model performance: comparison of area under the curve of different algorithms for each outcome.
Model | Outcome of interest | ||||
---|---|---|---|---|---|
EFSa | OSb | GFc | AGVHDd | CGVHDe | |
Random forest entire dataset | 0.7900 | 0.7925 | 0.8024 | 0.6793 | 0.7320 |
XGBoostf | 0.7754 | 0.7785 | 0.7948 | 0.6731 | 0.7230 |
Logistic regression | 0.7464 | 0.7835 | 0.7578 | 0.6925 | 0.7019 |
Naïve Bayes | 0.6930 | 0.7111 | 0.7107 | 0.6386 | 0.6384 |
Adaboost | 0.7452 | 0.7806 | 0.7561 | 0.6934 | 0.7005 |
Support vector classifier | 0.7357 | 0.7810 | 0.7561 | 0.6841 | 0.7061 |
EFS: event-free survival.
OS: overall survival.
GF: graft failure.
AGVHD: acute graft-versus-host disease.
CGVHD: chronic graft-versus-host disease.
XGBoost: extreme gradient boosting.