I am running a CART Decision Tree, with 5 independent variables and I wonder how can I interpret the results when the same variable splits the terminal node and also splits the terminal node closest parent. For instance..
The independent variable is Income and one terminal node says x <= 11.000 "Income" and its parent says x > 5.0000, so it seems that they overlap (between 5.000 and 11.000) which can be a way to interpret or on the other hand should I only consider the terminal node figure x <= 11.000?
The question is:
Node 18 should be “read” like the customers with income between 10.001and 15.000 OR just the customers with income above $15.000?