Table 6 Total predictor features and the results of the best-performing ML model for different coefficient thresholds.

From: Machine learning-based models for screening of anemia and leukemia using features of complete blood count reports

Used features

Total predictor features

Model

Accuracy

Recall

All variables

20

RF

0.87

0.66

≥ 0.1

15

RF

0.88

0.73

≥ 0.2

9

RF

0.82

0.58

≥ 0.3

4

RF

0.80

0.51

Back to article page