Table 6 Performance metrics for fine-tuned ResNet50 model.

From: Fine-tuned deep learning models for early detection and classification of kidney conditions in CT imaging

Epoch

Test accuracy (%)

Validation accuracy (%)

The absolute variance between test and validation accuracy

Time required for each step (ms/step)

1

89.82

99.80

9.98

32

5

99.55

99.72

0.17

24

10

99.49

99.96

0.47

24

15

99.59

99.84

0.25

24

20

99.49

99.88

0.39

24

25

99.85

99.92

0.07

24

Back to article page