Table 8 Four-classes analysis overall results: confusion matrix from ESHRE guidelines diagnostic work-up.

From: Machine Learning (ML) based-method applied in recurrent pregnancy loss (RPL) patients diagnostic work-up: a potential innovation in common clinical practice

  

Predicted class

1 st

2 nd

3 th

4th

Real Class

1 st

76.20 ± 1.04%

13.32 ± 0.70%

3.89 ± 0.56%

6.59 ± 0.82%

2 nd

15.92 ± 0.83%

52.19 ± 1.29%

10.40 ± 0.92%

20.48 ± 1.10%

3 th

11.00 ± 0.37%

18.66 ± 0.61%

44.43 ± 2.17%

25.91 ± 2.16%

4 th

6.74 ± 0.38%

13.53 ± 0.56%

19.50 ± 1.25%

60.22 ± 1.29%

  1. ACCub (unbalanced accuracy) = 58.52 ± 0.58%;
  2. ACCb (balanced accuracy) = 58.51 ± 0.58%.
Back to article page