TABLE 6. Performance of Two Class Problems Using Different Settings of Various Classifiers Except the Binary SVM Classifier.

Methods	KNN classifier, with Cityblock Distance			Ensemble of learners classifier with Discriminant Learner and ‘RUSBoost’ method			ANN classifier with N= 40 and training function ’trainscg’
Problem	Sen. (%)	Spe. (%)	Acc. (%)	Sen. (%)	Spe. (%)	Acc. (%)	Sen. (%)	Spe. (%)	Acc. (%)
Bleeding versus Normal	94.93	97.00	96.29	94.43	95.17	94.81	91.51	96.68	95.67
Ulcer versus Normal	85.23	95.33	93.14	84.99	92.43	90.24	72.81	96.76	93.24
Tumor versus Normal	71.11	94.21	91.74	71.27	91.01	89.17	79.15	97.04	92.31
Bleeding versus Ulcer	96.12	94.41	95.53	91.73	91.80	91.67	95.73	91.80	91.23
Bleeding versus Tumor	97.15	95.99	97.31	95.13	91.05	94.34	97.05	92.29	95.23
Ulcer versus Tumor	94.14	90.23	92.58	86.11	78.58	82.12	90.14	87.94	88.91
Disease versus Normal	90.15	90.21	90.19	85.16	88.23	88.09	80.43	91.79	86.98
Bleeding versus Other	91.23	97.12	91.16	90.13	94.14	93.12	86.81	97.07	94.17
Ulcer versus Other	85.01	96.12	95.12	83.25	91.22	90.01	72.52	97.01	93.54
Tumor versus Other	71.11	96.64	94.59	72.13	86.82	86.13	75.13	98.11	93.51