Skip to main content
. 2020 Jan 23;3:8. doi: 10.1038/s41746-019-0209-7

Fig. 4. Untrustworthy predictions in the GRACE risk model using the trust score.

Fig. 4

a Calibration curves, b normalized Brier scores, and c AUCs for the most untrustworthy predictions (red) and those that fall within the bottom 99% trust score values (black). Error bars are from 100 bootstrap splits and show one standard deviation (a) or one standard error of the mean (b and c). Error bars are truncated at 1.