Skip to main content

Table 6 Uncertainty quantification performance of the GBT models and the GP models

From: Development and evaluation of uncertainty quantifying machine learning models to predict piperacillin plasma concentrations in critically ill patients

Model

ADCE

DCE

Sharpness (std) (mg/L)

GUH evaluation: a priori

GBT new

0.06

0.01

23.48 (11.41)

GP new

0.29

0.29

41.22 (4.26)

GUH evaluation: a posteriori

GBT prev

0.07

0.04

17.98(8.62)

GP prev

0.28

0.28

28.94 (0.86)

UMCG evaluation: a priori

GBT new

0.62

0.62

25.50 (13.43)

GP new

0.39

− 0.39

42.75 (9.67)

UMCG evaluation: a posteriori

GBT prev

0.31

− 0.31

17.61 (8.02)

GP prev

0.15

0.08

28.22 (0.99)

  1. Bold indicates the best model for that metric and case