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

Table 2:

Performance metrics. SFM is the proposed model.

EHR FLCHAIN SUPPORT SEER SLEEP
Calibration slope

DATE 0.7537 0.9668 0.9068 0.9161 0.9454
DRAFT 3.2138 5.4183 2.9640 2.0763 25.2855
S-CRPS 1.6246 1.9662 1.1795 1.1613 2.5746
CPH 2.5543 1.9116 1.3909 1.4358 3.8278
SFM 0.7734 0.9807 0.9405 0.9540 1.0235

Mean CoV

DATE 0.2477 0.3585 0.2987 0.1485 0.5168
DRAFT 5.0305 6.2952 3.8689 3.4501 8.4918
S-CRPS 0.8585 0.9412 0.7351 0.6036 1.0240
CPH - - - - -
SFM 0.2953 0.4484 0.3930 0.1993 0.5045

C-Index

DATE 0.7756 0.8264 0.8421 0.8320 0.7416
DRAFT 0.7796 0.8341 0.8560 0.8310 0.7617
S-CRPS 0.7704 0.8286 0.8685 0.8298 0.7529
CPH 0.7542 0.8344 0.8389 0.8223 0.6435
SFM 0.7786 0.8318 0.8319 0.8314 0.7491