Variable Selection for Decision Trees (Uplift Modelling)
Apr 12, 2017 8:11 AM(4531 views)
I would like to know what method JMP is using for the variable selection in decision trees.
I could not find an answer to this question. I read in a paper by Radcliffe and Surry that Low Qini Estimates (LQE's) are their suggestion for it.
I am fairly new to JMP and definetly not a statistic expert.
I am currently writing my Bachelor thesis with a partner company, where I am supposed to measure the uplift of their marketing campaigns. I have to accumulate 40 pages on this, so just the results won't do it, I need to know say what and why things happen.
If anyone has experience with uplift modelling of marketing campaigns, but can't answer the question above, still let me know. Your experience could be of great help to me!