BookmarkSubscribeRSS Feed
JASilver
Calcite | Level 5

Hi,

I'm currently going through the Applied Analytics with EMiner course and am on the stepwise regression portion.

When I alter the default settings for the entry and stay significance level I get very different results than the demo. When I ran the regression with the default, my results matched those produced by the demo. Tried to attached a word file with print screen, but this wasn't allowed.

Best,

Josh Silverman

Message was edited by: Anna Brown Message was edited by: Anna Brown Welcome to the community, JASilver! I'm omitting your email address in your question for security reasons. You're in good hands here on the Data Mining Community - bunches of SAS Enterprise Miner experts are here to help and will respond on the community thread. Maybe you can try to save the printed screen image separately and upload it as an image into the message? Anna

4 REPLIES 4
JASilver
Calcite | Level 5

Inserting pictures. Here's the demonstration (left side of the picture). Right side of the picture is what I get when I change the entry and stay significance levels from the default using the PVA97NK dataset.

Optimizing Regression Complexity (26 or 46).png

For some reason I can't get the pictures of the demo (correct results to paste in this message. I'll try another message (nothing is simple).

JASilver
Calcite | Level 5

I cant' get pictures of the demo results to upload for some reason. They look like the legitimate results a stepwise regression (versus the results I generated). Thanks!

M_Maldonado
Barite | Level 11

How much did you change the significance level?

From your plot it looks like only 1 variable stays in your regression model. I doubt it is a good answer.

What happens when you use a Model Comparison node to compare the suggested settings of the book vs your tweaked regression? Which model is better?

Cynthia_sas
SAS Super FREQ

Hi:

  It is my understanding that when you start the training, there is a Learning Path, where you choose your version of Enterprise Miner. The instructions and demos that you get will then show demos and solutions that were run in THAT version of EM. So, for example, if you are watching the EM 6 training lectures and demos, but using EM 13, it was my understanding that you might get different results than the exercise or practice solutions. I don't know whether this is what you are running into. Did you follow the Learning Path instructions to select your version and did you also perform the required setup steps to get the data that was delivered with this version of the training? We do sometimes get these questions at training@sas.com where someone reports that they are getting different results than what is shown in the course and although it has (rarely) been an error in the course, most often when we track it down, the student is watching the lectures for a different version of EM than the one in which they are performing the practices. So they are "out of sync" and need to switch the course to their version of EM (or the closest version...for example, if you have EM 13, you would watch the EM 12 training lectures/exercises).

cynthia

sas-innovate-2024.png

Don't miss out on SAS Innovate - Register now for the FREE Livestream!

Can't make it to Vegas? No problem! Watch our general sessions LIVE or on-demand starting April 17th. Hear from SAS execs, best-selling author Adam Grant, Hot Ones host Sean Evans, top tech journalist Kara Swisher, AI expert Cassie Kozyrkov, and the mind-blowing dance crew iLuminate! Plus, get access to over 20 breakout sessions.

 

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
  • 4 replies
  • 1121 views
  • 0 likes
  • 3 in conversation