Table 1 Classification accuracies and the number of selected genes for the two different optimizationmethods (Greedy and Simulated Annealing (SA)). For Greedy selection, the accuracies as well as the number of genes selected in iterations 1 and 2 are shown. The reported accuracies are LOOCV accuracies while the number of genes is the smaller subset common to all LOOCV experiments.
From: An Entropy-based gene selection method for cancer classification using microarray data
Algorithm | Colon | Breast | SRBCT | |||
|---|---|---|---|---|---|---|
| Â | % Acc | # Genes | % Acc | # Genes | % Acc | # Genes |
Greedy | 90.3/91.9 | 29/9 | 89.8/89.8 | 31/12 | 100/100 | 58/14 |
SA | 87.1/- | 26/- | 89.8/- | 44/- | 100/- | 58/- |