Research Article
Integrated Use of Statistical-Based Approaches and Computational Intelligence Techniques for Tumors Classification Using Microarray
Table 4
ACRs for thirteen approaches using cross-validation.
| | Method | ACR | | AML | ALL | Full sample |
| | Single stage | | | | | FLDA | 37.50% | 93.75% | 75.00% | | LR | 62.50% | 87.50% | 79.17% | | ANN | 50.00% | 93.75% | 79.17% | | SVM | 25.00% | 93.75% | 70.83% | | MARS | 50.00% | 75.00% | 66.67% | | Two-stage | | | | | LR-FLDA | 62.50% | 81.25% | 75.00% | | LR-ANN | 50.00% | 93.75% | 79.17% | | LR-SVM | 75.00% | 81.25% | 79.17% | | LR-MARS | 62.50% | 93.75% | 83.33% | | MARS-FLDA | 75.00% | 75.00% | 75.00% | | MARS-LR | 75.00% | 75.00% | 75.00% | | MARS-ANN | 37.50% | 93.75% | 75.00% | | MARS-SVM | 62.50% | 87.50% | 79.17% |
|
|
