Figure 4From: ETHNOPRED: a novel machine learning method for accurate continental and sub-continental ancestry identification and population stratification correctionA comparison of 10-fold cross validation accuracy of individual decision trees and ensembles of disjoint decision trees of variable size in European population classification problem using HapMap phase III datasets. An ensemble of 3 disjoint decision trees involving 31 SNPs has a 10-fold cross validation accuracy of 86.5% ± 2.4% which is significantly better than the baseline accuracy of 61.8%.Back to article page