Figure 4

Total sample sizes needed for classification to majority poorly-separated HapMap populations CHB, JTP, CHD and GIH. For the linear classifier based on the SNP data from the three populations, the estimated learning curve gives the required total sample size for different values of the threshold, γ, satisfying $\widehat{\mathit{\text{VUS}}\left(\infty \right)}-\widehat{\mathit{\text{VUS}}\left(n\right)}<\gamma$. Here, ρ = 1, α = 0.1, m = 92, and $\widehat{\mathit{\text{VUS}}\left(\infty \right)}=0.5580$.