Skip to main content
. Author manuscript; available in PMC: 2022 Apr 29.
Published in final edited form as: IEEE Trans Neural Netw Learn Syst. 2023 Apr 4;34(4):1666–1680. doi: 10.1109/TNNLS.2020.3029631

Figure 2:

Figure 2:

Test set calibration and variation visualized for two datasets: seer and sleep (rows). Left: proportion of events of interest vs. predicted events. A perfectly calibrated model will follow the (dashed) diagonal line. Right: coefficient of variation (CoV) distributions. The legend shows the percentage of test set events covered by 95% intervals from predicted time-to-event distributions.