Skip to main content
. 2018 Apr 26;6(2):20. doi: 10.3390/proteomes6020020

Figure 2.

Figure 2

Graphical representation of the pre-treatment effects by data centering and autoscaling methods. This figure represents five samples, for which each of the vertical boxes is the feature value distribution of one sample, with the mean depicted as a horizontal bar inside the box. Centering removes the offset from the data so that the sample means become zero and autoscaling converts the data so that the standard deviation becomes one.