Table 4 Comparative analysis with baseline models.
From: A hybrid XAI-driven deep learning framework for robust GI tract disease diagnosis
Model | MCC | Train time (s) | Infer time (s) | Accuracy | Precision | Recall | F1-score |
|---|---|---|---|---|---|---|---|
Hybrid model (proposed model) | 0.9195 | 3676.63 | 0.1888 | 92.93% | 0.9327 | 0.9293 | 0.9296 |
Swin Transformer (ViT) | 0.0000 | 2109.64 | 0.0110 | 11.87% | 0.0141 | 0.1187 | 0.0252 |
ResNet (CNN) | 0.8998 | 1936.14 | 0.0066 | 91.16% | 0.9172 | 0.9116 | 0.9116 |
EfficientNet (CNN) | 0.8637 | 1937.42 | 0.0140 | 87.75% | 0.8989 | 0.8775 | 0.8759 |
