TABLE 3.

Prediction performance of the established logistic regression models.

LR model	Feature size	Cohort	AUC (95%CI)	ACC	SPE	SEN
LR-NSCLC	7	Training	0.860 (0.775–0.944)	0.796	0.925	0.725
		Validation	0.819 (0.665–0.972)	0.809	0.889	0.800
LR-BC	5	Training	0.909 (0.850–0.969)	0.839	0.925	0.825
		Validation	0.872 (0.771–0.973)	0.787	0.852	0.800
LR-EGFR	6	Training	0.850 (0.735–0.965)	0.700	0.550	1.000
		Validation	0.750 (0.514–0.987)	0.700	0.800	0.800
LR-HER2	5	Training	0.900 (0.800–1.000)	0.850	0.850	0.900
		Validation	0.830 (0.645–1.000)	0.750	0.800	0.800