Table 7.
Classification Results for Colon Data set
| Number of Genes | |||
| Ranking | 8 | 16 | 32 |
| Univariate (One-Dimensional) Ranking | |||
| P-value (P-val) | 84.4 | 88.8 | 88.6 |
| Fold-Change (FC) | 86.9 | 87.3 | 88.5 |
| Frequency (Freq) | 89.6 | 89.9 | 91.9 |
| Two-Dimensional Ranking | |||
| P-val & FC (1) | 88.7 | 88.8 | 90.0 |
| P-val & FC (2) | 88.7 | 88.7 | 88.8 |
| P-val & FC (3) | 86.6 | 88.7 | 88.7 |
| P-val & Freq (1) | 88.9 | 89.7 | 89.9 |
| P-val & Freq (2) | 89.1 | 86.7 | 90.5 |
| P-val & Freq (3) | 89.8 | 90.4 | 90.6 |
| FC & Freq (1) | 88.3 | 89.9 | 91.4 |
| FC & Freq (2) | 87.2 | 88.1 | 91.9 |
| FC & Freq (3) | 88.7 | 90.0 | 91.5 |
| Three-Dimensional Ranking | |||
| P-val & FC & Freq (1) | 86.3 | 89.0 | 90.3 |
| P-val & FC & Freq (2) | 88.4 | 89.1 | 89.8 |
| P-val & FC & Freq (3) | 88.7 | 86.7 | 90.5 |
Gene ID, p-value, fold-change, and the layer assigned by the (1) point-admissible, (2) convex, and (3) Pareto algorithms for 70 genes selected based on the p-value 0.001 and 2.0-fold as the cutoff from the 2,000 genes in the colon data set. The p-values are computed based on 50,000 permutations.