IBM SPSS Modeler is a data mining workbench that helps you build predictive models quickly and intuitively, without programming. Choose from two editions: SPSS Modeler Professional and SPSS Modeler Premium. These versions are available in both desktop and server configurations.
The following passage explains the concept of Combining Models in Modeler. If you would like to know more about training courses, visit the IBM Training Course Catalog.
Combining Models in Modeler
One of the major advantages of IBM SPSS Modeler is that it is very easy to build and combine models (combining models typically leads to increased predictive accuracy). Combining models in Modeler has become even easier with the Ensemble Node. To use this node, one would have to build at least two models and place these models in the same stream, as shown below. These models then have to be connected to the Ensemble Node.
There are many ways of combining models but one basic procedure uses the following logic:
Create two or more models.
When the models agree, use that prediction.
When the models don’t agree, use the model prediction with the highest confidence.
The setup for the above procedure is shown below (however the Ensemble Node is capable of combining models using other procedures).
In this way the Ensemble Node makes it extremely easy to combine models and thereby reap the benefits of such a procedure.
Follow @basupp on Twitter for updates, news, articles and videos related to Business Analytics Support.