Table 3 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 LEU dataset.
LEU 5-Fold | KNN | SVMs | ||||||
|---|---|---|---|---|---|---|---|---|
| Â | 30 | 60 | 100 | Best Accuracy/#Genes | 30 | 60 | 100 | Best Accuracy/#Genes |
GS2 | 0.961 ± 0.048 | 0.968 ± 0.044 | 0.971 ± 0.040 | 0.971/85 | 0.958 ± 0.052 | 0.967 ± 0.047 | 0.974 ± 0.039 | 0.974/98 |
GS1 | 0.965 ± 0.048 | 0.973 ± 0.040 | 0.979 ± 0.034 | 0.979/100 | 0.965 ± 0.050 | 0.970 ± 0.043 | 0.979 ± 0.037 | 0.979/93 |
Cho's | 0.958 ± 0.049 | 0.963 ± 0.046 | 0.968 ± 0.043 | 0.968/100 | 0.953 ± 0.054 | 0.962 ± 0.053 | 0.970 ± 0.043 | 0.970/98 |
F-test | 0.960 ± 0.049 | 0.966 ± 0.045 | 0.974 ± 0.038 | 0.974/96 | 0.957 ± 0.055 | 0.968 ± 0.049 | 0.975 ± 0.039 | 0.975/99 |
LEU LOO | KNN | SVMs | ||||||
| Â | 30 | 60 | 100 | Best Accuracy/#Genes | 30 | 60 | 100 | Best Accuracy/# Genes |
GS2 | 0.944 | 0.972 | 0.958 | 0.986/10 | 0.958 | 0.958 | 0.972 | 0.986/25 |
GS1 | 0.958 | 0.986 | 0.972 | 0.986/60 | 0.972 | 0.986 | 0.986 | 0.986/4 |
Cho's | 0.944 | 0.944 | 0.958 | 0.972/9 | 0.958 | 0.958 | 0.986 | 0.986/80 |
F-test | 0.944 | 0.944 | 0.972 | 0.986/25 | 0.958 | 0.958 | 0.972 | 0.986/33 |