Hi There,
How should i interpret the view of the decision learner. Let's take for example the first split. Wich of the leaves is the survivor and wich one is not? Is it correct that in every split there is one survivor and one side is not?
Hi Robbert,
there is some description in the node description, did you see this already?
I don't know the notation of survivor in the concept of learning decision trees.
I just made you an example, please see the attached image. The setosa is the majoriy class of the whole workflow (48/142) and using all data the distributions of the classes can be found in the table.
Below you see Petal.Width, this is the first split attribut. By splitting via this, we get the highest Information Gain. Afterwards we have two leaves, one pure setosa, the other contains of versicolor annd viginica, this would be split again.
Did this help?
Cheers, Iris