Table 6 Training and validation performance across epochs on ISIC 2017 Dataset.
From: AttenUNeT X with iterative feedback mechanisms for robust deep learning skin lesion segmentation
Epochs | Training ACC | Training loss | Val—ACC | Val—loss | Precision’s | Recall’s |
|---|---|---|---|---|---|---|
10 | 0.8441 | 0.3600 | 0.8452 | 0.3504 | 0.7695 | 0.6572 |
25 | 0.8473 | 0.3388 | 0.8556 | 0.3310 | 0.8173 | 0.6335 |
50 | 0.8713 | 0.8713 | 0.8595 | 0.3199 | 0.8058 | 0.7122 |
75 | 0.8890 | 0.2545 | 0.8665 | 0.3209 | 0.8230 | 0.7745 |
100 | 0.9066 | 0.2150 | 0.8675 | 0.3655 | 0.8331 | 0.8329 |
150 | 0.9550 | 0.1134 | 0.9021 | 0.3922 | 0.9457 | 0.8991 |
