Table 2 Percentage accuracy of 10-fold cross validation of classification methods for all genes
From: A comparative study of different machine learning methods on microarray gene expression data
Dataset | SVM | RBF Neural Nets | MLP Neural Nets | Bayesian | J48 Decision Tree | Random Forest | Id3 | Bagging |
|---|---|---|---|---|---|---|---|---|
1. Lymphoma (Devos et.al, 2002) | 96.0 | 84.0 | 68.0 | 88.0 | 64.0 | 76.0 | 48.0 | 52.0 |
2. Breast Cancer (Perou et. al, 2000) | 97.6 | 97.6 | 96.4 | 92.9 | 92.9 | 96.4 | 94.0 | 96.4 |
3. Colon Cancer (Alon et. al, 1999) | 95.6 | 91.1 | 91.1 | 93.3 | 91.1 | 80.0 | 88.9 | 93.3 |
4. Lung Cancer (Garber et. al, 2001) | 97.2 | 97.2 | 97.2 | 95.8 | 94.4 | 95.8 | 97.2 | 97.2 |
5. Adenocarcinoma (Beer et.al, 2002) | 96.5 | 94.2 | 75.6 | 75.6 | 74.4 | 79.1 | 66.3 | 79.1 |
6. Lymphoma (Alizadeh et al, 2000) | 96.9 | 88.5 | 75.0 | 85.4 | 75.0 | 76.0 | 62.5 | 84.4 |
7. Melanoma (Bittner et. al, 2000) | 94.7 | 81.6 | 84.2 | 76.3 | 81.6 | 81.6 | 52.6 | 81.6 |
8. Ovarian Cancer (Welsh et. al, 2001) | 94.9 | 84.6 | 89.7 | 87.2 | 87.2 | 89.7 | 74.4 | 89.7 |