Table 4 Test AUROCs for XGBoost for predicting 1 year mortality using all features, pre-transplant features only, and the LAS.

From: In silico perturbations provide multivariate interpretability in predicting post-lung transplant outcomes

All features

Pre-transplant features

LAS

0.63 [0.60, 0.67] 95% CI

0.60 [0.57,0.64] 95% CI

0.53 [0.51,0.55] 95% CI

Back to article page