I'm using Enterprise Miner for the first time and am trying to create a Decision Tree that will select records that have the highest propensity to become a conversion of a prospect to a customer. If I have a binary Yes/No target variable and I want a decision tree to show me the variable values that point to a Yes decision, how do I modify the Decision Tree so that the model will ultimately select those records that are most likely to me a Yes. Since most of my training data is a No, the tree seems to favor the decisions that lead to No. Does that make sense? Let me know if you need any clarification.
When you are fitting a Decision Tree model to a categorical outcome, the Tree will calculate probabilities for each possible outcome. You can change the level associated with the event by changing the sort order of the target variable but this will not yield a different model. You seem to be describing a rare event scenario where the event of interest happens far less frequently. There is another thread in this community re: your options in this scenario at
Hope this helps!
Doug
When you are fitting a Decision Tree model to a categorical outcome, the Tree will calculate probabilities for each possible outcome. You can change the level associated with the event by changing the sort order of the target variable but this will not yield a different model. You seem to be describing a rare event scenario where the event of interest happens far less frequently. There is another thread in this community re: your options in this scenario at
Hope this helps!
Doug
Available on demand!
Missed SAS Innovate Las Vegas? Watch all the action for free! View the keynotes, general sessions and 22 breakouts on demand.
Use this tutorial as a handy guide to weigh the pros and cons of these commonly used machine learning algorithms.
Find more tutorials on the SAS Users YouTube channel.