Dear All, @AlexanderFillbrunn
I have run the Decision Tree & i have exported the rules using write Csv.
My question is, when i exported i get the total count and number of correct count. But, number of correct count is not consistent; consistent in the sense it gives for some node count for zero and for some node it gives count of 1 and hence i can not identify number of records which one for zero count and which one is for 1 count (as it does not provide any column header)
Hence, i am unable to calculate probability of 1 that i am interested in.
How can i solve this problem? Or is there any way to get predicted probabilities for 1 directly while exporting the rule file?
Maybe take a look at this example
I have explored that, but that is not solving my problem mentioned above.
could you show us screenshots or upload your workflow? I don’t really understand your question. Which column contains 0? Is it “Number of correct”? That would be really strange, since the predicted class in a leaf node is determined by the majority class.
The probability of a single record is there along with a ruleset that gives you the information for every leaf of the tree how the probability was calculated using the absolute number of correct and not correct records.
Maybe you explain your question further or you could even provide us with a sample workflow that demonstrates your problem.
Hi @AlexanderFillbrunn, @mlauber71
I got the answer. Thank you very much for your help
This topic was automatically closed 182 days after the last reply. New replies are no longer allowed.