- when we are building a decisiontree, we are usually calculating the gini impurity at each node.
- I am interested to see the range of gini impurity in case of more than 2 classes.
- Because entropy always have range=[0,1], irrespective of number of classes.
I have seen everywhere gin impurity have range=[0, 0.5]. But it is not true in the case of more than 2-classes.