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

Table 6

Accuracies for test datasets using 281 features.

Datasets	Number of features	SVM			KNN			NB
		Acc	Sn	Sp	Acc	Sn	Sp	Acc	Sn	Sp
A1	281	78.57	80	42.86	57.14	60	28.57	57.14	40	57.14
A2	281	66.67	55.56	83.33	60.00	66.67	50	60.00	66.67	50
A3	281	60.00	66.67	50	40.00	33.33	50	66.67	77.78	50
T1	281	73.33	66.67	83.33	53.33	44.44	66.67	60.00	55.56	66.67
T2	281	81.25	77.78	85.71	68.75	66.67	71.43	87.50	77.78	100
T3	281	61.11	75	50	61.11	62.5	60	61.11	75	50
W1	281	87.50	88.89	85.71	75.00	55.56	100	93.75	100	85.71
W2	281	75.00	77.78	71.43	75.00	66.67	85.71	75.00	77.78	71.43
W3	281	88.24	100	75	82.35	77.78	87.5	88.24	100	75
A1, A2, and A3	281	86.67	57.14	100	72.41	42.86	81.82	55.17	28.57	63.64
T1, T2, and T3	281	60.00	57.14	60.71	65.71	57.14	67.86	57.14	57.14	57.14
W1, W2, and W3	281	70.59	42.86	77.78	85.29	42.86	96.3	79.41	14.29	96.3
All	281	86.30	28.57	92.42	83.56	14.29	90.91	63.01	57.14	63.64