Figure 3From: Breast cancer prediction using genome wide single nucleotide polymorphism dataAccuracy of the BestKNN algorithm for different numbers of MeanDiff selected SNPs. Accuracy of the classifiers built using BestKNN on sets of SNPs with the top {500, 600, ..., 1500} MeanDiff scores. This suggests that our model is fairly robust to the number of MeanDiff-selected SNPs, when selecting more than 500 SNPs.Back to article page