Table 1 Classification accuracy results obtained by the F-statistics measure for different feature subsets with different numbers of top ranked features (SNPs) in 30 simulations, on 3 and 4 populations, respectively
From: Effective selection of informative SNPs and classification on the HapMap genotype data
Feature Numbers | Mean accuracy ± std (minimal/maximal accuracy) (%) for 3 populations | Mean accuracy ± std (minimal/maximal accuracy) (%) for 4 populations |
|---|---|---|
1 | 69.21 ± 1.60 (64.29/70) | 54.98 ± 1.60 (51.43/57.14) |
10 | 72.96 ± 7.82 (64.29/92.86) | 56.16 ± 2.36 (45.71/58.57) |
20 | 74.48 ± 7.82 (65.71/95.71) | 57.88 ± 4.26 (54.29/74.29) |
30 | 74.92 ± 8.79 (65.71/95.71) | 58.47 ± 5.18 (48.57/74.29) |
40 | 77.29 ± 10.55 (65.71/97.14) | 59.51 ± 5.08 (54.29/77.14) |
50 | 79.75 ± 11.96 (64.29/98.57) | 61.18 ± 7.24 (54.29/82.86) |
60 | 82.46 ± 11.41 (67.14/98.57) | 64.09 ± 7.10 (57.14/82.86) |
70 | 84.68 ± 11.15 (67.14/98.57) | 64.48 ± 7.73 (57.14/82.86) |
80 | 94.48 ± 7.03 (64.29/98.57) | 67.98 ± 8.86 (55.71/84.29) |
90 | 93.74 ± 5.98 (68.57/98.57) | 70.84 ± 9.13 (57.14/87.14) |
100 | 93.79 ± 3.44 (80/98.57) | 73.99 ± 7.09 (58.57/87.14) |