TABLE 2.

Performance comparison of various ML techniques on the full features for prediction of soybean charcoal rot disease.

Method	TP	FP	TN	FN	Accuracy	Sensitivity	Specificity	Precision	NPV	F1 score	MCC	AUC
LR-L1	1153	47	1149	51	95.92 ± 8.32	95.75 ± 8.74	96.08 ± 7.93	96.01 ± 8.05	95.84 ± 8.60	95.88 ± 8.39	91.84 ± 16.63	97.37 ± 5.82
LR-L2	1143	57	1151	49	95.58 ± 9.00	95.92 ± 8.90	95.25 ± 9.16	95.31 ± 9.11	95.87 ± 8.90	95.61 ± 8.99	91.18 ± 17.99	97.05 ± 6.58
MLP	1139	61	1138	62	94.88 ± 9.58	94.83 ± 10.66	94.92 ± 8.51	94.72 ± 8.97	95.06 ± 10.16	94.77 ± 9.83	89.76 ± 19.14	96.69 ± 7.39
RF	1143	57	1147	53	95.42 ± 9.64	95.58 ± 9.27	95.25 ± 10.01	95.34 ± 9.80	95.50 ± 9.47	95.46 ± 9.54	90.83 ± 19.28	97.20 ± 6.35
GBT	1168	32	1155	45	96.79 ± 6.49	96.25 ± 7.88	97.33 ± 5.16	97.16 ± 5.55	96.49 ± 7.29	96.68 ± 6.75	93.62 ± 12.90	98.42 ± 3.42
SVM	1150	50	1155	45	96.04 ± 7.55	96.25 ± 7.88	95.83 ± 7.26	95.81 ± 7.36	96.29 ± 7.76	96.03 ± 7.61	92.09 ± 15.10	97.86 ± 4.65