Table 5.

Phase-II (nCOVID vs. Pneumonia) classification performance of different supervised models using Training-Testing set in 10-fold cross-validation setup.

Classification algorithms	Accuracy (±STD)	Specificity (±STD)	Precision (±STD)	Recall (±STD)	F1-Measure (±STD)	AUC (±STD)	MCC (±STD)
Without using Augmented Images
DT	72.98 ± 3.54	70.93 ± 12.06	73.06 ± 6.20	74.96 ± 8.82	73.41 ± 3.22	0.73 ± 0.04	0.47 ± 0.07
ANN	74.29 ± 5.28	74.29 ± 32.99	74.29 ± 8.57	74.29 ± 30.07	74.29 ± 17.5	0.74 ± 0.05	0.49 ± 0.13
SVM (RBF Kernel)	80.03 ± 4.40	80.18 ± 10.61	81.05 ± 8.12	79.91 ± 7.44	80.01 ± 4.30	0.80 ± 0.04	0.61 ± 0.09
KNN	80.17 ± 4.62	83.30 ± 7.82	82.68 ± 6.42	77.03 ± 7.23	79.47 ± 4.92	0.80 ± 0.05	0.61 ± 0.09
NB	80.46 ± 6.03	79.57 ± 6.26	80.00 ± 5.71	81.36 ± 7.70	80.58 ± 6.18	0.80 ± 0.06	0.61 ± 0.12
SVM (Poly Kernel)	83.34 ± 3.46	85.89 ± 5.63	85.42 ± 4.37	80.77 ± 5.81	82.87 ± 3.63	0.83 ± 0.03	0.67 ± 0.07
SVM (Linear Kernel)	85.07 ± 6.94	87.09 ± 9.64	87.11 ± 8.73	83.08 ± 7.39	84.83 ± 6.74	0.85 ± 0.07	0.71 ± 0.14
Using Augmented Images
NB	84.63 ± 6.75	85.94 ± 6.26	85.51 ± 6.43	83.31 ± 8.46	84.32 ± 7.12	0.85 ± 0.07	0.69 ± 0.13
DT	84.84 ± 5.27	84.19 ± 4.82	84.36 ± 4.85	85.49 ± 6	84.91 ± 5.34	0.85 ± 0.05	0.7 ± 0.11
ANN	94.93 ± 2.54	94.2 ± 3.52	94.29 ± 3.1	95.65 ± 3.5	94.96 ± 2.55	0.95 ± 0.03	0.9 ± 0.05
KNN	97.27 ± 1.86	95.83 ± 2.5	95.98 ± 2.35	98.7 ± 1.25	97.31 ± 1.81	0.97 ± 0.02	0.95 ± 0.04
SVM (RBF Kernel)	97.7 ± 1.85	97.41 ± 2.17	97.49 ± 2.27	97.99 ± 1.91	97.71 ± 1.82	0.98 ± 0.02	0.95 ± 0.04
SVM (Poly Kernel)	97.98 ± 1.99	97.7 ± 2.26	97.75 ± 2.18	98.19 ± 1.58	97.89 ± 2.05	0.98 ± 0.02	0.95 ± 0.04
SVM (Linear Kernel)	98.78 ± 0.96	98.14 ± 1.69	98.19 ± 1.68	99.23 ± 0.72	98.79 ± 0.95	0.99 ± 0.01	0.98 ± 0.02