Table 2 CNN and CFNN Adam’s optimizer hyperparameters that achieved lowest RMSE after grid-searching
Hyperparameter | Initial Learning Rate (η) | Beta 1 (β 1) | Beta 2 (β 2) | epsilon ϵ |
|---|---|---|---|---|
CNN (CT Scans) | 0.001 | 0.9 | 0.999 | 10−8 |
CFNN (Clinical Factors) | 0.01 | 0.99 | 0.999 | 10−8 |