BMC Bioinformatics

Table 6 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 DLBCL dataset.

From: A stable gene selection in microarray data analysis

DLBCL 5-Fold	KNN				SVMs
	30	60	100	Best Accuracy/# Genes	30	60	100	Best Accuracy/# Genes
GS2	0.881 ± 0.081	0.906 ± 0.078	0.914 ± 0.074	0.916/98	0.872 ± 0.081	0.918 ± 0.068	0.933 ± 0.054	0.933/98
GS1	0.878 ± 0.078	0.895 ± 0.075	0.903 ± 0.076	0.903/100	0.861 ± 0.079	0.895 ± 0.075	0.917 ± 0.066	0.918/98
Cho's	0.874 ± 0.085	0.909 ± 0.075	0.920 ± 0.072	0.920/99	0.869 ± 0.085	0.915 ± 0.068	0.930 ± 0.061	0.930/99
F-test	0.877 ± 0.079	0.893 ± 0.078	0.902 ± 0.080	0.902/100	0.869 ± 0.079	0.910 ± 0.074	0.925 ± 0.063	0.926/95
DLBCL LOO	KNN				SVMs
	30	60	100	Best Accuracy/# Genes	30	60	100	Best Accuracy/# Genes
GS2	0.883	0.922	0.922	0.935/70	0.896	0.961	0.948	0.961/55
GS1	0.896	0.896	0.922	0.922/85	0.870	0.883	0.961	0.961/81
Cho's	0.883	0.922	0.922	0.935/69	0.909	0.896	0.948	0.961/74
F-test	0.896	0.896	0.883	0.922/61	0.857	0.935	0.948	0.961/92

Back to article page

ISSN: 1471-2105

Contact us

General enquiries: journalsubmissions@springernature.com