Table 4 Two-classes analysis overall results. Model Trained on the 18 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

2-classes analysis

Predicted class

Healthy

Affected (RPL)

Real class

Healthy

88.98 ± 0.66%

11.01 ± 0.66%

Affected (RPL)

1.29 ± 0.13%

98.71 ± 0.13%

  1. ACCub (unbalanced accuracy) = 96.28 ± 0.19%;
  2. ACCb (balanced accuracy) = 93.85 ± 0.34%.
Back to article page