Table 5 Performance of the proposed models (Xception and Inception-ResNetV2) on kaggle and Harvard dataverse datasets.
From: Trustworthy deep learning for malaria diagnosis using explainable artificial intelligence
Model | Dataset | Validation accuracy (%) | Test accuracy (%) |
|---|---|---|---|
Xception | Kaggle dataset | 97.7 | 97.6 |
Inception-ResNetV2 | Kaggle dataset | 97.8 | 97.8 |
Xception | Harvard dataverse | 97.7 | 97.5 |
Inception-ResNetV2 | Harvard dataverse | 97.6 | 97.6 |
