Table 5 Accuracy comparison of disease prediction methods using feature selection (FS) methods.
Dataset | Methods | SelectKBest | Lasso | Recursive feature elimination(RFE) | CBOA |
|---|---|---|---|---|---|
Heart disease | FedHFP+RNN | 80.31 | 82.45 | 83.33 | 85.15 |
FedHFP+LSTM | 81.78 | 83.64 | 85.39 | 86.79 | |
DeFedHDP+EDBN | 84.25 | 85.80 | 87.03 | 88.78 | |
FedAvgBC+TabNet | 85.87 | 87.56 | 89.21 | 91.41 | |
PPFBXAIO+EDBN | 87.79 | 89.39 | 90.74 | 93.07 | |
PPFILBOXAI+EDBN | 89.87 | 91.66 | 92.79 | 95.71 | |
Breast cancer wisconsin | FedHFP+RNN | 81.56 | 83.54 | 84.31 | 85.58 |
FedHFP+LSTM | 82.95 | 84.81 | 85.96 | 87.17 | |
DeFedHDP+EDBN | 84.38 | 86.72 | 87.90 | 89.28 | |
FedAvgBC+TabNet | 86.62 | 88.25 | 89.83 | 92.44 | |
PPFBXAIO+EDBN | 88.77 | 90.50 | 91.58 | 95.07 | |
PPFILBOXAI+EDBN | 91.05 | 92.36 | 93.75 | 96.84 |
