Fig. 4: Machine learning: binary and ordinal classification maps.
From: A worldwide study of subcortical shape as a marker for clinical staging in Parkinson’s disease
Binary and ordinal classification uses vertex-wise thickness information from all subcortical structures. The color bars represent the learned weights of the classification model, positive values (SD from the learned weights) in red, and negative values in blue. More intense colors indicate a stronger predictive power of the classification. Displayed are the results of A the binary classification of people with PD and controls, B the ordinal classification of HY1–HY2–HY345.
