Hi, I am having problems with the IGN Node of SAS Eminer. I grouped my data outside Eminer using MS Excel until I arrived to my desired WOE trend for each variable. I then noted of the significant IV's. However when I entered the groupings in SAS Eminer (v5.2) and run the IGN Node, the result showed that some of my variables have zero IV's. And all of these variables are categorical/nominal. Can you help me identify what went wrong?
nsdominguez, thank you for your question. We are digging into the details of this scenario using SAS Enterprise Miner (v5.2) and will get back to you as soon as possible with input. Stay tuned.
Hello - in the case of a value of zero for a variable's IV, the procedure could not find any splits for that variable. In such a case, all values are in the same group. There are a number of possibilities for this, including too many missing values or restrictions you imposed such as the minimum number of observations required to be a separate bin. The answer is really data-dependent and would have to be answered on a case-by-case basis. Have you tried skipping the use of Excel altogether and doing the data grouping in the IGN in Enterprise Miner?
Good news: We've extended SAS Hackathon registration until Sept. 12, so you still have time to be part of our biggest event yet – our five-year anniversary!
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.