Skip to main content

Advertisement

Table 8 Test accuracy results (mean ± std-dev%) of random forest models against the number of genes per leaf n min on the ten gene data sets.

From: Genome-wide association data classification and SNPs selection using two-stage quality-based Random Forests

Data set Model 1 genes 2 genes 5 genes 8 genes 10 genes 15 genes X s X w
COL RF .844 ± 0.5 .818 ± 0.8 .832 ± 0.7 .830 ± 0.6 .849 ± 0.3 .853 ± 0.4   
  GRRF .865 ± 0.5 .832 ± 0.6 .848 ± 0.5 .838 ± 0.6 .853 ± 0.3 .859 ± 0.3   
  wsRF .845 ± 0.5 .837 ± 0.4 .857 ± 0.5 .834 ± 0.6 .844 ± 0.4 .848 ± 0.5   
  ts-RF .877 ± 0.4 .863 ± 0.4 .879 ± 0.3 .863 ± 0.5 .874 ± 0.3 .874 ± 0.3 245 317
SRB RF .959 ± 0.3 .957 ± 0.2 .961 ± 0.2 .944 ± 0.5 .914 ± 1.0 .777 ± 1.2   
  GRRF .976 ± 0.2 .972 ± 0.1 .972 ± 0.2 .941 ± 0.7 .898 ± 1.1 .802 ± 1.1   
  wsRF .968 ± 0.3 .967 ± 0.3 .966 ± 0.3 .957 ± 0.3 .912 ± 0.5 .771 ± 0.2   
  ts-RF .977 ± 0.2 .974 ± 0.1 .977 ± 0.1 .962 ± 0.4 .922 ± 1.1 .812 ± 1.1 606 546
LEU RF .826 ± 1.2 .849 ± 0.9 .866 ± 0.9 .879 ± 0.9 .871 ± 1.0 .874 ± 1.0   
  GRRF .873 ± 0.9 .867 ± 0.7 .880 ± 0.9 .878 ± 0.9 .876 ± 0.9 .885 ± 0.9   
  wsRF .848 ± 1.0 .848 ± 0.9 .863 ± 1.0 .858 ± 1.1 .851 ± 1.0 .866 ± 1.1   
  ts-RF .893 ± 0.7 .885 ± 0.6 .906 ± 0.7 .908 ± 0.7 .913 ± 0.7 .905 ± 0.7 502 200
LYM RF .972 ± 0.2 .983 ± 0.1 .979 ± 0.3 .930 ± 1.1 .855 ± 1.2 .823 ± 0.6   
  GRRF .991 ± 0.1 .989 ± 0.1 .983 ± 0.3 .928 ± 1.1 .840 ± 1.1 .805 ± 0.4   
  wsRF .981 ± 0.2 .982 ± 0.2 .975 ± 0.4 .928 ± 0.2 .845 ± 0.3 .801 ± 0.2   
  ts-RF .993 ± 0.1 .995 ± 0.0 .987 ± 0.3 .935 ± 1.1 .856 ± 1.2 .828 ± 0.7 1404 275
BR2 RF .627 ± 0.7 .618 ± 0.7 .608 ± 0.7 .622 ± 0.7 .601 ± 0.7 .640 ± 0.7   
  GRRF .713 ± 0.9 .623 ± 0.8 .615 ± 0.8 .627 ± 0.7 .617 ± 0.8 .643 ± 0.7   
  wsRF .634 ± 0.7 .627 ± 0.8 .618 ± 0.8 .619 ± 0.9 .604 ± 0.8 .626 ± 0.7   
  ts-RF .788 ± 0.7 .766 ± 0.8 .776 ± 0.9 .776 ± 0.8 .765 ± 1.1 .780 ± 0.8 194 631
BR3 RF .560 ± 0.7 .568 ± 0.7 .560 ± 0.7 .581 ± 0.6 .563 ± 0.8 .567 ± 0.8   
  GRRF .635 ± 0.8 .580 ± 0.6 .574 ± 0.7 .586 ± 0.6 .568 ± 0.7 .580 ± 0.8   
  wsRF .572 ± 0.7 .575 ± 0.7 .571 ± 0.7 .579 ± 0.4 .565 ± 0.8 .580 ± 0.6   
  ts-RF .654 ± 0.7 .657 ± 0.7 .661 ± 0.6 .670 ± 0.6 .645 ± 0.7 .648 ± 0.9 724 533
NCI RF .589 ± 1.1 .584 ± 1.3 .558 ± 1.2 .470 ± 1.2 .379 ± 1.5 .206 ± 0.9   
  GRRF .631 ± 1.3 .592 ± 1.3 .561 ± 1.2 .483 ± 1.2 .403 ± 1.5 .228 ± 1.0   
  wsRF .594 ± 1.1 .589 ± 1.4 .578 ± 1.0 .478 ± 1.2 .390 ± 1.5 .239 ± 1.4   
  ts-RF .742 ± 1.2 .731 ± 1.8 .684 ± 1.3 .552 ± 1.9 .430 ± 1.7 .248 ± 1.1 247 1345
BRA RF .708 ± 1.6 .706 ± 2.0 .701 ± 1.7 .637 ± 2.1 .600 ± 3.0 .368 ± 3.0   
  GRRF .748 ± 1.7 .729 ± 1.9 .726 ± 1.8 .654 ± 2.3 .650 ± 4.0 .416 ± 2.9   
  wsRF .708 ± 1.8 .718 ± 1.9 .691 ± 1.8 .652 ± 1.7 .650 ± 3.2 .431 ± 2.3   
  ts-RF .819 ± 1.6 .815 ± 2.0 .783 ± 1.8 .694 ± 2.1 .679 ± 3.0 .405 ± 3.4 1270 1219
PRO RF .887 ± 0.4 .894 ± 0.4 .895 ± 0.4 .891 ± 0.3 .882 ± 0.3 .891 ± 0.3   
  GRRF .929 ± 0.2 .916 ± 0.2 .916 ± 0.2 .908 ± 0.3 .907 ± 0.3 .917 ± 0.2   
  wsRF .908 ± 0.2 .911 ± 0.3 .913 ± 0.3 .906 ± 0.3 .897 ± 0.3 .908 ± 0.3   
  ts-RF .926 ± 0.2 .928 ± 0.1 .927 ± 0.2 .919 ± 0.2 .915 ± 0.2 .926 ± 0.2 601 323
ADE RF .840 ± 0.4 .846 ± 0.4 .849 ± 0.3 .845 ± 0.4 .832 ± 0.3 .839 ± 0.3   
  GRRF .855 ± 0.5 .842 ± 0.4 .848 ± 0.3 .848 ± 0.4 .832 ± 0.3 .834 ± 0.4   
  wsRF .841 ± 0.4 .841 ± 0.4 .845 ± 0.3 .842 ± 0.4 .828 ± 0.3 .832 ± 0.3   
  ts-RF .909 ± 0.4 .906 ± 0.4 .904 ± 0.4 .902 ± 0.5 .888 ± 0.4 .901 ± 0.4 108 669
  1. Numbers in bold are the best results.