Skip to main content

Table 3 The average 10-fold cross validation accuracy of ensembles of size m, for m = 1..30 in continental population classification problem

From: ETHNOPRED: a novel machine learning method for accurate continental and sub-continental ancestry identification and population stratification correction

Number of models in the ensemble (m)

Number of ensembles

The average 10-fold cross validation accuracy of ensembles of size m (Acc)

1

30

95.38

2

435

91.34

3

4060

98.36

4

27405

97.03

5

142506

99.32

6

593775

98.81

7

2035800

99.67

8

5852926

99.44

9

14300000

99.93

10

30000000

99.92

11

54600000

99.98

12

86500000

99.96

13

120000000

99.99

14

145000000

99.99

15

155000000

99.99

16

145000000

99.99

17

120000000

99.99

18

86500000

99.99

19

54600000

99.99

20

30000000

99.99

21

14300000

99.99

22

5852926

99.99

23

2035800

99.99

24

593775

99.99

25

142506

99.99

26

27405

99.99

27

4060

99.99

28

435

99.99

29

30

99.99

30

1

100