Table 2

10-fold cross-validation resultsa of the SVM with different kernel functions

Kernel function

Sensitivity

Specificity

Accuracy


SVM-Linear

84.03 ± 0.74

74.90 ± 1.47

79.46 ± 0.87

SVM-Polynomial

84.56 ± 1.06

76.38 ± 2.14

80.47 ± 0.87

SVM-RBF

85.91 ± 2.60

80.54 ± 2.64

83.22 ± 1.76


a SVMs with different kernel functions were trained with training set (1).

Li and Lai BMC Bioinformatics 2007 8:353   doi:10.1186/1471-2105-8-353

Open Data