Sorting Rules from a Decision Tree
based on the Titanic tutorial data I am using the decision tree model to analyse those data. I know that in this case the outcoming tree is really clear, but I was wondering how it is, if the tree might become bigger. Using the Operator "Tree to Rules" I was able to get out the rules for each path of the tree. Behind each rule is written in brackets, how many People with exactly those characteristics survived, for example (3 / 25) - what means 12%. Is it possible to sort all the rules by the chance of surviving, given the defined characteristics?
Thanks to one post from mschmitz ("How to Get Rules from a Decision Tree?"), I am able to classify the rules by the label "survived" (yes/no). With modifying this script I wasn't successful. I guess my main problem is, that all attributes in this script - including numbers - are defined as String. Just changing the datatype doesn't work.
Is there any other possibility how to sort those rules, generated by "Tree to Rules"?
I am thankful for every Kind of help