Figure 11.

The result of classification using a decision tree created for 10 features. Classification errors are marked using squares. For this rotation of coordinate system the issues related to the common part of both separated groups are roughly visible. When selecting other “ w” features in the visualisation, this boundary is even less visible. Looking at these examples of the features w(1), w(2) and w(3) on each axis, a possible separation of healthy subjects from patients is visible. For other features, as shown further on, these results are much better. For a full visualisation of all the features, a 10-dimensional space should be used.