Skip to main content

Table 4 Selected features \(i^{p}\), \(i^{ps}\), \(i^{n}\), \(i^{ns}\) are existing positive, selected positive, existing negative and selected negative features for each labelling intervals, respectively

From: Deep self-supervised machine learning algorithms with a novel feature elimination and selection approaches for blood test-based multi-dimensional health risks classification

Intervals

\(i^{p}\)

\(i^{ps}\)

\(i^{n}\)

\(i^{ns}\)

 ± 100%

1

1

1.303

300

 ± 90%

81

81

16.380

300

 ± 80%

11

11

4.877

300

 ± 70%

19

19

2.048

300

 ± 60%

34

34

604

30

 ± 50%

60

60

213

213

 ± 40%

75

75

101

101

 ± 30%

5

5

35

35

 ± 20%

0

0

19

19

 ± 10%

1

1

1

1