RIDDLE: Race and ethnicity Imputation from Disease history with Deep LEarning
Fig 2
Receiver operating characteristic (ROC) curves.
ROC curves and their corresponding area under the curve (AUC) values were calculated for each of the four race and ethnicity classes. Micro-average (all cases considered as binary, e.g., Hispanic vs. non-Hispanic) and macro-average (average across classes) curves were also computed. Data and metrics for a representative experiment is shown. Across experiments, the mean micro-average AUC was 0.874, and the macro-average AUC was 0.833.