Figure 2
From: Problems with the nested granularity of feature domains in bioinformatics: the eXtasy case
Features of the synthetic data. Kernel density estimates of the synthetic data distributions. Panels in the first row display distributions of three non-informative features for two classes separately. Note that decreased density in the vicinity of sampling interval borders is in fact an artifact of the kernel density estimation procedure and not of the uniform sampling itself, thus it affects only this visualization. The bottom three panels depict informative features.