Table 3.

LUAD and LUSC Classification Statistics.

Feature selection method	Accuracy	Specificity	Sensitivity	Precision	F-measure	95% Bootstrap confidence interval
DGE (Top 500)	0.932476	0.901235	0.966443	0.9	0.932039	(0.9035, 0.9614)
PCA (Top 500)	0.942122	0.901235	0.986577	0.90184	0.942308	(0.9132, 0.9678)
mRMR (Top 500)	0.916399	0.888889	0.946309	0.886792	0.915584	(0.8842, 0.9453)
Lasso (68 Genes)	0.938907	0.907407	0.973154	0.90625	0.938511	(0.9100, 0.9646)
Xgboost (148 Genes)	0.935691	0.901235	0.973154	0.900621	0.935484	(0.9068, 0.9614)
Overlapping 131 Genes	0.938907	0.895062	0.986577	0.896341	0.939297	(0.9100, 0.9646)
17 Proposed Biomarkers	0.92926	0.889	0.9735	0.88957	0.9296	( 0.9003, 0.9550 )