Table 1 Comparison of classification performance with three types of markers in the breast cancer dataset
Classifiers | Markers | Sensitivity | Specificity | Accuracy |
|---|---|---|---|---|
kNN | genes | 93.6 (2.06) | 82.56 (3.29) | 88.08 (1.94) |
pathways | 92.44 (1.08) | 89.53 (1.76) | 90.99 (1.2) | |
pathway-sets | 97.09 (2.06) | 90.7 (1.24) | 93.9 (1.03) | |
RF | genes | 87.21 (1.76) | 87.79 (1.08) | 87.5 (0.82) |
pathways | 93.31 (2.62) | 92.73 (1.49) | 93.02 (1.08) | |
pathway-sets | 93.31 (2.62) | 92.73 (1.49) | 93.31 (1.03) | |
SVM | genes | 86.63 (1.08) | 86.63 (2.41) | 86.63 (1.52) |
pathways | 93.9 (1.73) | 91.28 (1.08) | 92.59 (1.23) | |
pathway-sets | 91.57 (3.27) | 92.44 (1.08) | 92.01 (1.91) | |
Naïve Bayes | genes | 86.63 (1.08) | 86.63 (2.41) | 86.63 (1.52) |
pathways | 93.9 (1.73) | 91.28 (1.08) | 92.59 (1.23) | |
pathway-sets | 91.57 (3.27) | 92.44 (1.08) | 92.01 (1.91) |