cancel
Turn on suggestions
Auto-suggest helps you quickly narrow down your search results by suggesting possible matches as you type.
Showing results for 
Show  only  | Search instead for 
Did you mean: 
Try the Materials Informatics Toolkit, which is designed to easily handle SMILES data. This and other helpful add-ins are available in the JMP® Marketplace
Quick Start: Predictive Models - Tips & Tricks to Build Models Quickly - Predictive Modeling
Created: May 15, 2020 12:43 PM  |  Last Modified: May 22, 2020 2:12 PM

(view in My Videos)

 

Sr. Systems Engineer Kemal Oflus gives you helpful hints for building your models more effectively. In this presentation, Kemal offers guidance on a number of critical predictive modeling considerations, from improving your data cleanup and visualization to identifying the best predictive modeling option for your needs. You’ll also learn why Kemel recommends, before doing any model building, that you first visualize your data to better understand what’s in there. He also shows you how to use JMP Graph Builder to see if there’s anything hidden in your data that might help or impede your model building.

 

Comments
PatrickGiuliano
‎08-24-2020 05:18 PM

Hi @kemal_oflus, Thanks for your contribution!  I have a few perhaps alarmingly simple questions but I trust that your reply should be useful to the community.     

 

1.  Why is RSM appropriate for fitting a model on your predictor = Yield? 

2.  Is there a way in JMP to impose a predictor-constraint on Yield in this case, not less than 0% and not more than 100%, as a column property so that the appropriate model/models can be fit without the Yield > 100% problem that you highlighted in your example in Fit Model?

3. Why does the Classification (or Regression)-tree based approach impose a physical practical constraint on the Yield response? 

 

Cheers, @PatrickGiuliano 

Cheers, Patrick