turn on suggestions

Auto-suggest helps you quickly narrow down your search results by suggesting possible matches as you type.

Showing results for

- JMP User Community
- :
- Discussions
- :
- Discussions
- :
- Predictive Modeling Data Preparation (Inputs/Outputs Variables)

Topic Options

- Subscribe to RSS Feed
- Mark Topic as New
- Mark Topic as Read
- Float this Topic for Current User
- Bookmark
- Subscribe
- Printer Friendly Page

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

Jul 29, 2015 8:28 AM
(2989 views)

How do you set up your variables, such that, they appear as in JMP Pro 12 as inputs and outputs in the process of developing a predictive models? Some of the inputs variables include, but are not limited to the following: orders historical data, page views data, quotes data, demographic data, overlay data, abandoned cart data, email campaign data files, mail data, etc. Any piece of advice?

Jenkins Macedo

1 ACCEPTED SOLUTION

Accepted Solutions

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

Jul 29, 2015 1:33 PM
(5101 views)

If I understand your question correctly, it seems that you need to define what you are trying to predict vs. what inputs you will use to predict the outputs. Think of it this way: Y = f(X1, X2, X3, X4, ....Xn). Y=Predicted outputs X's = Input variables. Each column can be one Y or one X. I suggest you label them so that you understand what they represent rather than something like "Var_AX456", which might be a data aquisition system tag of some sort. Maybe it is "Reactor Temp.", which makes more sense.

Remember that data within a column must be consistent as discrete, continuous, text, etc. Make sure each column is give the correct characteristics (Data tType and Modeling Type).

Hope this helps.

Steve

3 REPLIES

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

Jul 29, 2015 1:33 PM
(5102 views)

If I understand your question correctly, it seems that you need to define what you are trying to predict vs. what inputs you will use to predict the outputs. Think of it this way: Y = f(X1, X2, X3, X4, ....Xn). Y=Predicted outputs X's = Input variables. Each column can be one Y or one X. I suggest you label them so that you understand what they represent rather than something like "Var_AX456", which might be a data aquisition system tag of some sort. Maybe it is "Reactor Temp.", which makes more sense.

Remember that data within a column must be consistent as discrete, continuous, text, etc. Make sure each column is give the correct characteristics (Data tType and Modeling Type).

Hope this helps.

Steve

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

Jul 30, 2015 6:53 AM
(2854 views)

Thanks smoore2. Your descriptions and explanation is exactly what I had thought about doing, but wasn't so sure.Thanks for the excellent feedback. Do you know of any resource in terms of a good textbook which explains how to develop predictive models using JMP Pro 12?

Jenkins Macedo

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Get Direct Link
- Email to a Friend
- Report Inappropriate Content

Jul 30, 2015 8:53 AM
(2854 views)

Macedo, There are several great resources. First and foremost, you can look at the books in the JMP Help menu. Click "Help" at the top of your screen. Then click "Books", then "Specialized Models". there you will find great techniques like Partitioning and Neural Networks. A great book is "Fundamentals of Predictive Analytics with JMP" by Klimberg and McCullough. Hope this helps!

Steve