Mean (± SD.) prediction performance, averaged over 10 independent experiments.

D-calibration counts the number of times the respective model was D-calibrated. Harrell’s CI, Uno’s CI and IBS results are multiplied by 100.D校准（D-calibration）用于统计对应模型被执行D校准的次数。Harrell一致性指数（Harrell’s CI）、Uno一致性指数（Uno’s CI）以及积分Brier得分（IBS）的结果均乘以100。