Research Article
Hybrid Disease Diagnosis Using Multiobjective Optimization with Evolutionary Parameter Optimization
Table 34
Parameters to be used in MLP for all datasets.
| Dataset | Learning rate | Momentum | Hybrid MLP accuracy | Base paper accuracy |
| Cleveland | 0.4410246832765716 | 0.945131728055943 | 85.8 | 85 | Statlog | 0.24687115044065697 | 0.7512112614957723 | 85.9 | 84.4 | Spect | 0.001620407197768992 | 0.5458467309532906 | 85 | 82.02 | Spectf | 0.0037254964036241137 | 0.6064034495456784 | 82.4 | 78.28 | Eric | 0.6953073769599724 | 0.9167657941184544 | 81.3 | 80.86 | Breast cancer | 0.5272364248697747 | 0.9288899224295802 | 96.99 | 96.71 | Hepatitis | 0.6376255545427609 | 0.9250563419048221 | 87.1 | 86.45 | Thyroid | 0.1516498076389815 | 0.48805304429332785 | 97.7 | 94.81 | Parkinson | 0.8486064853474067 | 0.3499016503919223 | 93.8 | 89.23 | Pima Indian diabetics | 0.03218577681226653 | 0.06466339445401592 | 77.60 | 77.08 | BUPA | 0.8329619224653821 | 0.014749643317800043 | 73 | 67.54 |
|
|