Table 3.

Hybrid (Haralick + (GLCM) Features based detection performance and utilizing ML Algorithms to differentiate NSCLC from SCLC.

Methods	Sensitivity	Specificity	PPV	NPV	Accuracy	FPR	AUC
Naïve Bayes (NB)	0.9449	0.9326	0.9568	0.9146	0.9402	0.06742	0.9895
SVM RBF	1	0.9831	0.9895	1	0.9935	0.01685	1
Decision Tree (DT)	0.9858	0.9775	0.9858	0.9775	0.9826	0.02247	0.9895
SVM Gaussian	1	0.9888	0.9929	1	0.9956	0.01124	1
SVM Polynomial	0.9982	0.9972	0.9982	0.9972	0.9978	0.002809	0.9995