Journal of Clinical Bioinformatics

Table 3 Overall accuracy in total, training and validation sets by SVM #2 method

From: Using gene expression data to identify certain gastro-intestinal diseases

TOTAL SET				TRAINING SET				VALIDATION SET
Tc*	Ti^†	TOTAL^‡	%I^§	Tc	Ti	TOTAL	%I	Tc	Ti	TOTAL	%I
80% IBS-C (RBF kernel)
152	8	160	5	124	3	127	2	28	4	33	12
80% IBD-C (RBF kernel)
207	2	209	1	160	0	166	0	41	2	43	4
80% IBD-IBS (RBF kernel)
139	4	143	3	111	1	113	1	27	3	30	10
60% CD-UC (RBF kernel)
77	7	85	9	47	4	51	8	31	3	34	11
60% IBS-C (polynomial)
150	10	160	6	91	4	95	4	59	6	65	9
60% IBD-C (polynomial)
195	14	209	7	88	7	95	7	107	7	114	6
60% IBD-IBS (polynomial)
124	19	143	13	78	8	85	8	46	11	58	19
60% CD-UC (polynomial)
76	9	85	10	47	4	50	8	30	5	35	14

*Tc, total number correct in designated set.
^†Ti, total number incorrect in designated set.
^‡Total, total number of cases and controls analyzed in designated set.
^§% I, incorrect percentage of case:control calls in designated set.

Back to article page