Table 4 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 GLIOMA dataset.

From: A stable gene selection in microarray data analysis

GLIOMA 5-Fold	KNN				SVMs
	30	60	100	Best Accuracy/# Genes	30	60	100	Best Accuracy/# Genes
GS2	0.660 ± 0.141	0.670 ± 0.140	0.671 ± 0.140	0.676/90	0.651 ± 0.134	0.679 ± 0.133	0.699 ± 0.131	0.701/97
GS1	0.674 ± 0.143	0.677 ± 0.148	0.679 ± 0.141	0.684/97	0.659 ± 0.145	0.698 ± 0.137	0.720 ± 0.138	0.722/99
Cho's	0.659 ± 0.145	0.660 ± 0.141	0.652 ± 0.131	0.664/31	0.618 ± 0.141	0.662 ± 0.131	0.668 ± 0.132	0.670/96
F-test	0.647 ± 0.140	0.663 ± 0.142	0.667 ± 0.133	0.674/91	0.639 ± 0.138	0.672 ± 0.131	0.684 ± 0.130	0.685/84
GLIOMA LOO	KNN				SVMs
	30	60	100	Best Accuracy/#Genes	30	60	100	Best Accuracy/# Genes
GS2	0.760	0.700	0.660	0.780/28	0.700	0.660	0.760	0.760/96
GS1	0.700	0.760	0.740	0.780/35	0.680	0.700	0.760	0.760/45
Cho's	0.720	0.640	0.640	0.820/20	0.640	0.680	0.620	0.720/2
F-test	0.700	0.660	0.700	0.780/70	0.640	0.620	0.740	0.740/100

Back to article page