BookmarkSubscribeRSS Feed
🔒 This topic is solved and locked. Need further help from the community? Please sign in and ask a new question.
GalacticAbacus
Obsidian | Level 7

Hi,

 

I'm interested in comparing the Transformation node vs. the Interactive Binning Node.

 

What is the difference between setting the interval-inputs setting to "optimal binning" for a variable vs. using the interactive binning node?

 

Let's say I'm using the Transformation node and I've selected the interval-inputs to "optimal", and based on chi-squared a couple of my variables have been selected for binning transformation. Is there a benefit to tagging on the Interactive Binning Node downstream to assess optimal binning once transformation is assessed?

 

I'm assuming it's still of use as some of my variables transformed via log or non-binning transformation could still be considered downstream for binning once normality is addressed... so binning is only redundant if a variable was binned via transformation and then re-binned with the binnning node?

 

I'd appreciate clarification of any misunderstanding I may have.

1 ACCEPTED SOLUTION

Accepted Solutions
WendyCzika
SAS Employee

Those 2 nodes just give you different methods for binning, so you would just use one of those nodes for binning, based on whatever method you want to use.  Optimal binning in Transform node takes the target into account to come up with the best bins, while in Interactive Binning, your options are to do either bucket or quantile binning (which don't use the target), then you can interactively modify those bins.  Once the variables are binned, via either node, they are no longer interval, so you wouldn't bin those again, but as you said, you could still bin a variable after it's been log-transformed for example and is still interval.  Hope that helps!

View solution in original post

3 REPLIES 3
WendyCzika
SAS Employee

Those 2 nodes just give you different methods for binning, so you would just use one of those nodes for binning, based on whatever method you want to use.  Optimal binning in Transform node takes the target into account to come up with the best bins, while in Interactive Binning, your options are to do either bucket or quantile binning (which don't use the target), then you can interactively modify those bins.  Once the variables are binned, via either node, they are no longer interval, so you wouldn't bin those again, but as you said, you could still bin a variable after it's been log-transformed for example and is still interval.  Hope that helps!

GalacticAbacus
Obsidian | Level 7

Yes it does help, thank you much!

AnshulS
Obsidian | Level 7

Thank you for your reply, which one will you prefer to use for Credit Risk - Probability of Default Model (if I do not have Interactive Grouping node in my SAS EM tool.

sas-innovate-2024.png

Available on demand!

Missed SAS Innovate Las Vegas? Watch all the action for free! View the keynotes, general sessions and 22 breakouts on demand.

 

Register now!

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
  • 3 replies
  • 4200 views
  • 1 like
  • 3 in conversation