. 2024 Mar 14;12:e17098. doi: 10.7717/peerj.17098

Table 4. Predictive performance of different models on training and validation sets.

Model	AUC (SD)	Cutoff (SD)	Accuracy (SD)	Sensitivity (SD)	Specificity (SD)	F1 score (SD)
Training set
XGBoost	0.913(0.002)	0.200(0.025)	0.810(0.015)	0.858(0.023)	0.800(0.022)	0.600(0.012)
LR	0.739(0.001)	0.162(0.002)	0.625(0.009)	0.757(0.013)	0.599(0.013)	0.402(0.001)
LightGBM	0.738(0.001)	0.167(0.000)	0.834(0.000)	0.726(0.003)	0.750(0.001)	NaN
AdaBoost	0.805(0.001)	0.397(0.005)	0.725(0.006)	0.782(0.009)	0.711(0.008)	0.485(0.003)
GNB	0.794(0.001)	0.134(0.003)	0.738(0.002)	0.744(0.005)	0.737(0.003)	0.486(0.001)
MLP	0.768(0.022)	0.170(0.008)	0.697(0.039)	0.735(0.016)	0.689(0.048)	0.448(0.026)
SVM	0.508(0.059)	0.180(0.044)	0.607(0.189)	0.411(0.306)	0.646(0.287)	0.205(0.116)
Validation set
XGBoost	0.860(0.008)	0.200(0.025)	0.781(0.021)	0.837(0.046)	0.742(0.041)	0.556(0.026)
LR	0.738(0.009)	0.162(0.002)	0.625(0.009)	0.766(0.045)	0.601(0.046)	0.402(0.010)
LightGBM	0.738(0.010)	0.167(0.000)	0.834(0.000)	0.726(0.024)	0.750(0.013)	NaN
AdaBoost	0.804(0.010)	0.397(0.005)	0.723(0.012)	0.764(0.050)	0.736(0.044)	0.481(0.017)
GNB	0.793(0.010)	0.134(0.003)	0.738(0.010)	0.748(0.037)	0.740(0.025)	0.487(0.014)
MLP	0.767(0.020)	0.170(0.008)	0.696(0.039)	0.736(0.043)	0.698(0.041)	0.449(0.034)
SVM	0.501(0.057)	0.180(0.044)	0.603(0.191)	0.486(0.297)	0.578(0.287)	0.217(0.094)

Notes.