Table 5 Four-classes analysis overall results. Model Trained on the all the available features selected in leave 200 out cross validation over 1000 iterations: Confusion matrix.

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

0 abortion

1 abortion

2–3 abortions

≥4 abortions

Real class

0 abortion

91.46 ± 0.33%

8.05 ± 0.21%

0.47 ± 0.25%

0.02 ± 0.06%

1 abortion

0.93 ± 0.23%

89.39 ± 0.95%

7.59 ± 0.83%

2.08 ± 0.45%

2–3 abortions

3.26 ± 0.22%

11.82 ± 0.50%

68.86 ± 0.76%

16.06 ± 0.69%

≥4 abortions

2.55 ± 0.07%

8.67 ± 0.36%

11.17 ± 0.85%

77.62 ± 0.91%

  1. ACCub (unbalanced accuracy) = 81.86 ± 0.35%;
  2. ACCb (balanced accuracy) = 81.83 ± 0.35%.
Back to article page