Table 3

Accuracy percentiles and characteristics of models from 50 cross-validation samples of DS1, m/z > 1500

Algorithm

Test Set Accuracy 25th, 75th Percentile

Median # Clusters

Median # Markers

Proportion of Perfect Chromosomes


GA(p1 = 0, p2 = 0)

.83, .87

90

11

1.0

GA(0, .001)

.80, .87

93

11

1.0

GA(0, .002)

.82, .87

92.5

9

1.0

GA(.002, 0)

.87, .92

12

19

.26

GA(.002, .001)

.90, .93

14

13

.38

GA(.002, .002)

.88, .92

13

7.5

.20

GA(.005, 0)

.87, .92

5

20

.02

GA(.005, .001)

.87, .90

5

10

.02

GA(.005, .002)

.87, .92

5

6.5

0

GA(.008, 0)

.85, .90

3.5

18.5

0

GA(.008, .001)

.85, .89

4

8.5

0

GA(.008, .002)

.85, .90

4

6

0

Best GA

.87, .91

7

10

.12

Boosting

.93, .97

NA

NA

NA

PAM

.80, .84

NA

NA

NA


Jeffries BMC Bioinformatics 2004 5:180   doi:10.1186/1471-2105-5-180

Open Data