From: An alternative approach to dimension reduction for pareto distributed data: a case study
Categorical variable
Theil’s U
A/label
0.13
B/label
0.09
C/label
D/label
0.03