Figure 6
From: Missing value imputation improves clustering and interpretation of gene expression microarray data
Effect of random initialization of k -means. Summary of the results when initializing the k-means clustering from 10 random initializations (k = 5). (A) Agreement with the original clusters (see the legend of Fig. 3). (B) Agreement with the original GO terms (see the legend of Fig. 4). The error bars represent the standard error of mean (SEM) values over the 10 random cluster initializations.