BMC Bioinformatics

Table 7 The leave-one-out and 5-fold cross validation classification accuracies of the SVM-classifier and the KNN-classifier based on four gene selection methods, GS1, GS2, Cho's, and F-test, on the MLL dataset.

From: A stable gene selection in microarray data analysis

MLL 5-Fold	KNN				SVMs
	30	60	100	Best Accuracy/#Genes	30	60	100	Best Accuracy/#Genes
GS2	0.937 ± 0.056	0.947 ± 0.055	0.948 ± 0.053	0.949/91	0.926 ± 0.058	0.941 ± 0.052	0.947 ± 0.051	0.947/87
GS1	0.946 ± 0.054	0.940 ± 0.057	0.942 ± 0.058	0.948/29	0.932 ± 0.059	0.947 ± 0.053	0.952 ± 0.050	0.952/99
Cho's	0.950 ± 0.048	0.954 ± 0.048	0.960 ± 0.045	0.960/93	0.942 ± 0.051	0.946 ± 0.050	0.955 ± 0.048	0.955/89
F-test	0.949 ± 0.050	0.950 ± 0.050	0.953 ± 0.051	0.954/99	0.943 ± 0.051	0.945 ± 0.053	0.948 ± 0.051	0.948/100
MLL LOO	KNN				SVMs
	30	60	100	Best Accuracy/# Genes	30	60	100	Best Accuracy/# Genes
GS2	0.944	0.958	0.972	0.972/90	0.917	0.958	0.944	0.972/91
GS1	0.958	0.944	0.958	0.972/97	0.958	0.958	0.958	0.972/56
Cho's	0.944	0.944	0.958	0.972/23	0.944	0.931	0.944	0.958/44
F-test	0.944	0.944	0.958	0.958/65	0.944	0.931	0.944	0.958/31

Back to article page

ISSN: 1471-2105

Contact us

General enquiries: journalsubmissions@springernature.com