Table 7 Comparison of proposed model.
From: Minimal sourced and lightweight federated transfer learning models for skin cancer detection
References | Model used | Accuracy |
|---|---|---|
“Convolutional neural network (ConvNet) model” | 86.90% | |
Pre-trained VGG16 and CNN | 78% | |
“Multi-layered CNN” | 93.58% | |
“CNN and SVM” | 85% | |
“MobileNet model” | 80.81% | |
“Multiple transfer learning models using deep learning” | 83.2% | |
Proposed model | Proposed DL on FL ecosystem (IID and None-IID model) | With IID 89.83% and Non-IID 90.64% |
