Skip to main content

Table 1 Classification of phenotypes against genes

From: Mimvec: a deep learning approach for analyzing the human phenome

 

AUC (%)

ACC (%)

BER (%)

 

LR

RF

SVM

LR

RF

SVM

LR

RF

SVM

Mimvec (50)

99.31

99.20

99.34

97.52

96.49

97.80

2.97

4.29

2.74

Mimvec (100)

99.41

98.92

99.31

97.83

96.17

97.78

2.60

5.19

2.75

Mimvec (150)

99.35

98.79

99.34

97.73

95.21

97.69

2.71

6.41

2.83

Mimvec (200)

99.43

98.67

99.26

97.75

94.43

97.67

2.65

7.99

2.85

Mimvec (250)

99.42

98.71

99.33

97.73

94.04

97.86

2.75

8.55

2.64

Mimvec (300)

99.36

98.50

99.17

97.70

93.59

97.60

2.75

9.32

2.91

  1. Disease and gene records can be well distinguished by vector representations of the records
  2. Bold numbers highlight performance achieved at the default setting (100 dimensions)
Back to article page

BMC Systems Biology

ISSN: 1752-0509

Contact us