Let's create an SPSS modeler flow by performing the following steps:
- To create an SPSS modeler flow, first we must go to Add to project:
- We can then use the Assets tab (as shown in the following screenshot) and click on the Modeler Flow icon:
- Next, type a name and description for the flow and select the IBM SPSS Modeler runtime (on the lower left) then click on the Create button (on the lower right):
- Now we are ready to create our machine learning modeler flow using the following flow editor canvas:
The flow editor palette empowers you to use machine learning, artificial intelligence (AI), and statistics modeling methods to derive new information from your data without the need for programming.
- Now we are ready to add our data, so we can drag a Data Asset node (found under Import) onto the canvas:
- Once you have a Data Asset node, you can double-click it, select Change data asset, select your preferred (training) file, and then click on OK. If you now right-click on the node, you can select Preview (shown in the following screenshot) to see your data (in read-only mode):
- To generate a profile of the data, you can add the Data Audit node (found under Outputs) by dragging it onto the canvas, connecting it to the Data Asset node and then clicking the VCR-type run icon:
You can see that the Data Audit node shows 25 Fields that it automatically found in the Data Asset node. Again, if you double-click on the Data Audit node, you'll be able to view and update various parameters, such as a name for the node (you can have multiple Data Audit nodes in a flow), as well as what to include in the output summarizations (graphs, basic and/or advanced statistics, and so on):
The outputs from the Data Audit node provides a detailed profile of your data. You can change parameters, rerun your flow, and recheck the results to become comfortable with the outputs:
The bottom page of the preceding screenshot is as follows:
