Skip to main content
. 2012 Sep 6;28(21):2824–2833. doi: 10.1093/bioinformatics/bts509

Fig. 6.

Fig. 6.

Reproducibility index for the real datasets, two classification rules (LDA and RBF-SVM) and two error estimation methods (5F-CV and LOO). The training sample size is 60, and d = 5 features are selected using t-test feature selection method: (a) LDA, 5F-CV; (b) RBF-SVM, 5F-CV; (c) LDA, LOO; (d) RBF-SVM, LOO