BookmarkSubscribeRSS Feed
NicolasC
Fluorite | Level 6

Hi. I was wondering whether there is a way in SAS Miner to group certain levels of ordinal variables. If one of my input variable is, for instance, "Number of Employee" and is stored as in 1-4, 5-9, 10-14 etc bins, is there a way to have Miner somehow group some of these levels, so that to have 1-9, 10-20. Even better, is there a way to have levels changed by Miner to find bins that would have better statistical significance (like 1-7, 7-18 etc...) in predicting the target (binary in this case)? For now, I do it ahead of my model in Guide but the choice of new levels is entirely based on my magic’s skill:) Many thanks. Nicolas

1 REPLY 1
PaigeMiller
Diamond | Level 26

I don't know the answer.

 

I do know that almost any type of predictive modeling will work better if you use the un-binned data instead of the binned data. Just something to thing about.

--
Paige Miller

Ready to join fellow brilliant minds for the SAS Hackathon?

Build your skills. Make connections. Enjoy creative freedom. Maybe change the world. Registration is now open through August 30th. Visit the SAS Hackathon homepage.

Register today!
How to choose a machine learning algorithm

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.

Discussion stats
  • 1 reply
  • 925 views
  • 0 likes
  • 2 in conversation