Skip to main content
. Author manuscript; available in PMC: 2023 Mar 15.
Published in final edited form as: Expert Syst Appl. 2022 Nov 6;214:119171. doi: 10.1016/j.eswa.2022.119171

Table 2.

Model performance in the hold-out test set and configuration parameters.

AUROC ACC Recall Spec F1 AP No. Features (1, 2, 3 grams)
GOS 0.94 [0.93–0.95] 0.77 [0.75–0.80] 0.77 [0.75–0.80] 0.92 [0.92–0.93] 0.77 [0.75–0.80] 0.65 [0.62–0.69] 243 (166, 65, 12)
mRS 0.90 [0.89–0.91] 0.59 [0.567–0.62] 0.59 [0.57–0.62] 0.93 [0.93–0.94] 0.59 [0.57–0.62] 0.41 [0.38–0.44] 536 (397, 116, 23)

The bootstrapping results in 95% confidence intervals are in parenthesis. AUROC – Area under the receiver operating characteristic curve, ACC – accuracy, Spec – specificity, AP – average precision, GOS – Glasgow Outcome Scale, mRS – modified Rankin Scale, C – inverse of regularization strength. For both outcome models, the configuration parameters selected for C and warm start were 0.05 and “True”, respectively.