Figure 1
From: Building gene expression profile classifiers with a simple and efficient rejection option in R
Probability density plots. Probability density plots to show how the value of class probability estimates of target and reject samples distribute over a set of experiments. In the MAX plot the considered random variable is the highest class probability estimate of each classified sample, split into target samples (solid line) and reject samples (dashed line). In the DIFF plot the considered random variable is the difference between the two highest probability estimates of each sample, considering target samples (solid line) and reject samples (dashed line).