BookmarkSubscribeRSS Feed
🔒 This topic is solved and locked. Need further help from the community? Please sign in and ask a new question.
chris_sonne
Calcite | Level 5

Hi everyone,

 

I'm using the SAS Enterprise Miner 14.1 to model and to assess the incremental response of a marketing activity. Following the successful application of the Incremental Response node, I'm trying to use the Model Comparison node. However, when appying the Model Comparison node as illustrated in the Getting Started with SAS EM Guide, I receive the following error message: "Must use one target variable modeled by a predecessor node".

 

Below, you can see a screenshot of my EM diagram. I only use one target (binary, response) and the role of this variable is specified before applying the incremental response nodes.

 

Does anyone know why I receive this error message and how to apply model comparison with two or more incremental response models properly?

 

 

Thanks in advance,

Christoph

 

 

EM_Model_Comparison.PNG

1 ACCEPTED SOLUTION

Accepted Solutions
Urban_Science
Quartz | Level 8

From the model Comparison Node help document:

Note: The Model Comparison node is designed to compare the performance of competing SAS Enterprise Miner process flow diagrams that use one (or more) analytic modeling nodes from the Model tab of the SAS Enterprise Miner toolbar. The Model Comparison node cannot compare the performance of process flow diagrams that do not use a model node from Model tab of the toolbar. For example, you cannot use the Model Comparison node to compare process flow diagrams that use a Survival node (from the Applications tab of the SAS Enterprise Miner toolbar) to make predictions. Survival node predictions (the probability that a target level will occur over a future time interval) are mathematically different in structure from modeling node predictions (the posterior probability that a target event will occur).

Since the Incremental Response node is not from the Model tab of the toolbar, the Model Comparison node cannot be used to compare models.

View solution in original post

3 REPLIES 3
Urban_Science
Quartz | Level 8

From the model Comparison Node help document:

Note: The Model Comparison node is designed to compare the performance of competing SAS Enterprise Miner process flow diagrams that use one (or more) analytic modeling nodes from the Model tab of the SAS Enterprise Miner toolbar. The Model Comparison node cannot compare the performance of process flow diagrams that do not use a model node from Model tab of the toolbar. For example, you cannot use the Model Comparison node to compare process flow diagrams that use a Survival node (from the Applications tab of the SAS Enterprise Miner toolbar) to make predictions. Survival node predictions (the probability that a target level will occur over a future time interval) are mathematically different in structure from modeling node predictions (the posterior probability that a target event will occur).

Since the Incremental Response node is not from the Model tab of the toolbar, the Model Comparison node cannot be used to compare models.

chris_sonne
Calcite | Level 5

Thank you for your answer.

 

It would be a nice feature of the SAS Enterprise Miner to show me a helpful error message when trying to use Model Comparison with the Incremental Response node.

AnnaBrown
Community Manager

Thanks for the feedback, @chris_sonne! Feel free to submit a SASWare Ballot Idea for this feature request so it can be considered for development.

 

Best,

Anna

SAS Innovate 2025: Call for Content

Are you ready for the spotlight? We're accepting content ideas for SAS Innovate 2025 to be held May 6-9 in Orlando, FL. The call is open until September 16. Read more here about why you should contribute and what is in it for you!

Submit your idea!

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
  • 2275 views
  • 0 likes
  • 3 in conversation