Figure 7
From: A highly efficient multi-core algorithm for clustering extremely large datasets
Clustering and stability estimation for HapMap SNP profiles. Cluster number estimation via repeated clustering of profiles/subjects for the HapMap data (210 profiles, 116678 SNPs). For each k ∈ {2, 3, ..., 10}, 1000 repeated cluster runs were performed. For each cluster number, two boxplots are shown, one summarizing the MCA values from the pairwise comparisons of all cluster results (left), and the other one for the results from the random prototype baseline (right). A higher value indicates increased stability.