Table 6 Empirical accuracy vs. Complexity Trade-off.
Dataset | Best Accuracy (%) | Method (from ensemble) | Std Dev | Ensemble Gain | Model complexity |
|---|---|---|---|---|---|
Satimage | 90.71 | Proposed 4 | 0.4688 | High | Moderate (36 features) |
EmailSpamdexing | 95.33 | Proposed 1,5,7,11 | ~ 1.0–2.5 | Very High | Medium |
Breast Cancer | 95.4 | Proposed 3 | 2.07 | Very High | High (due to large SD) |
Musk | 89.64 | Cosine | 2.47 | Medium | Moderate |
Iris | 97.71 | Proposed 11 | 1.54 | High | Low (4 features) |
