Fig 1.
Schematic of machine learning benchmarking study design.
Table 1.
Baseline characteristics of heart transplant recipients from October 18, 2018 to June 3, 2021.
Fig 2.
Violin plots comparing cross-validation C-indices for 7 different machine learning survival models against Cox PH in the A) post-policy cohort and B) pre-policy cohort.
Black bar indicates the mean c-index.
Table 2.
C-indices for holdout data.
Fig 3.
Average number of times variable is selected across all sparse machine learning models in the A) post-policy cohort and B) pre-policy cohort.
Sparse machine learning models include Lasso, Elastic Net, and Cox Boost.
Fig 4.
Top 20 significant predictor variables.
A) RSF in post-policy cohort B) RSF in pre-policy cohort C) Cox Boost in post-policy cohort D) Cox Boost in pre-policy cohort E) XGBT in post-policy cohort F) XGBT in pre-policy cohort.