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
- :
- Neural net model

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
- Permalink
- Email to a Friend
- Report Inappropriate Content

Aug 4, 2011 11:26 PM
(982 views)

Hi All,

I try to predict my data by use neural net model as output (Y) and variable input(X1,X2,X3,X4,X5) with 2 nodes. I'd like to find out the contributor (X) to prediction equation.

How do I analyse which input X have effect on Y? Thank you.

3 REPLIES

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

Aug 9, 2011 2:17 PM
(904 views)

I think that's one downside of neural networks: it's hard to interpret the input effects because of all the weigthed interconnections. I suppose you try Analysis > Modeling > Screening on the predicted values from the neural net model to give you some idea.

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

Aug 10, 2011 1:36 AM
(904 views)

Thank a lot for your suggestion Xan.

- Mark as New
- Bookmark
- Subscribe
- Subscribe to RSS Feed
- Permalink
- Email to a Friend
- Report Inappropriate Content

Aug 10, 2011 7:28 AM
(904 views)

In the Neural platform, you should be able to use the Prediction Profiler to explore which of your X's have large contributions to the variation in your output variable.

Michael Crotty

Sr Statistical Writer

JMP Development

Sr Statistical Writer

JMP Development