Table 3 Stage 1: F1-score, accuracy, precision and recall on the validation set for different models, K values and sampling methods (uni = uniform; sim = similarity). In bold, the best value for each metric. Top-left: F1-score. Top-right: Accuracy. Bottom-left: Precision. Bottom-right: Recall.

From: Classification of red cell dynamics with convolutional and recurrent neural networks: a sickle cell disease case study

  

A: CNN

B: C-RNN

sampling

sampling

uni.

sim.

uni.

sim.

K

10

0.48

0.47

0.56

0.55

20

0.54

0.55

0.56

0.56

30

0.50

0.51

0.58

0.58

50

0.52

0.53

0.61

0.61

K

10

0.92

0.89

0.95

0.95

20

0.99

0.99

0.95

0.95

30

0.93

0.94

0.95

0.95

50

0.94

0.94

0.96

0.96

K

10

0.32

0.30

0.39

0.38

20

0.36

0.38

0.40

0.39

30

0.31

0.33

0.42

0.42

50

0.33

0.35

0.46

0.46

K

10

0.93

0.90

0.97

0.96

20

0.97

0.98

0.96

0.95

30

0.92

0.94

0.95

0.95

50

0.94

0.95

0.94

0.93