A Statistical-Textural-Features Based Approach for Classification of Solid Drugs Using Surface Microscopic Images

Table 8

Accuracies for test datasets using top 2 features.

Datasets	Number of features	Feature name	SVM			KNN			NB
			Acc	Sn	Sp	Acc	Sn	Sp	Acc	Sn	Sp
A1	2	A189 and A226	50	30	57.14	57.14	70	14.29	50	50	28.57
A2	2	A189 and A226	53.33	55.56	50	66.67	66.67	66.67	60	55.56	66.67
A3	2	A189 and A226	53.33	55.56	50	53.33	66.67	33.33	46.67	33.33	66.67
T1	2	A189 and A226	80	77.78	83.33	73.33	88.89	50	60	44.44	83.33
T2	2	A189 and A226	68.75	44.44	100	62.5	55.56	71.43	50	44.44	57.14
T3	2	A189 and A226	55.56	75	40	55.56	75	40	55.56	62.5	50
W1	2	A189 and A226	75	66.67	85.71	75	100	42.86	62.5	66.67	57.14
W2	2	A189 and A226	81.25	88.89	71.43	56.25	66.67	42.86	68.75	66.67	71.43
W3	2	A189 and A226	88.24	88.89	87.5	64.71	55.56	75	76.47	55.56	100
A1, A2, and A3	2	A189 and A226	75.86	14.29	95.45	72.41	42.86	81.82	79.31	14.29	100
T1, T2, and T3	2	A189 and A226	60	71.43	57.14	60	42.86	64.29	71.43	14.29	85.71
W1, W2, and W3	2	A189 and A226	82.35	57.14	88.89	50	42.86	51.85	64.71	28.57	74.07
All	2	A189 and A226	73.97	57.14	75.76	82.19	28.57	87.88	90.41	14.29	98.48