Table 9 Evaluation results of FL clients and globally FL rounds.
From: Privacy preserving skin cancer diagnosis through federated deep learning and explainable AI
Client 1 | Client 2 | Client 3 | Globally FL | |||||
|---|---|---|---|---|---|---|---|---|
Training | Testing | Training | Testing | Training | Testing | Training | Testing | |
Accuracy | 96.57% | 86.34% | 96.70% | 88.52% | 100% | 87.43% | 95.65% | 86.52% |
Precision | 95.73% | 83.81% | 96.15% | 87.10% | 100% | 82.42% | 94.54% | 76.45% |
Recall | 97.55% | 91.67% | 97.14% | 92.04% | 100% | 90.36% | 96.37% | 90.69% |
Specificity | 95.56% | 80.46% | 95.63% | 87.37% | 100% | 84.00% | 95.00% | 84.17% |
Miss classification | 3.43% | 13.66% | 3.57% | 10.38% | 0% | 13.11% | 4.35% | 13.48% |
F1 score | 0.9664 | 0.8799 | 0.9664 | 0.8950 | 1.0 | 0.8620 | 0.9544 | 0.8293 |
