Shown below are balanced error rates


Number of significant SNPs
Nearest means
1 nearest neighbor
Logistic regression
Support vector machine
Support vector machine (degree 2 kernel)
Support vector machine (degree 3 kernel)
50 (Edward)
32.2 24.3 22.3 22.8
100 (Ittehad)
200 (Pratyusha)
32.2 25.3 25 24.2 24.1
300 (Kevin)
23.6
400 (Siddharth)
24.9
500 (Dima)
23.2 22.5 21.6 21.1
600 (Shijie)
34.6 23.4 22.2 20.3 20.5
700 (Roshan)
34.8 33.9 22.8 22.2 21.4 19.9
800 (Sruti)
35.5 24 23 21.5 21.1
900
1000