Building flows

A flow is a series of nodes that you connect on the canvas.

Flow

A flow is a group of data-processing operations that are connected in sequences. Flows represent the flow of data through each operation. Data flows from the data source through the sequence of operations to the end. Usually, a flow ends in a model or type of data output, such as a table or chart.

Flows are created by adding nodes on to the canvas and connecting them.

Canvas

The canvas is the main work area in SPSS Modeler, and it is where you build your flows.

Nodes

A node is a modular, self-contained set of operations. Nodes are a graphical way of representing these operations, and each node has a unique icon. These nodes can be linked together on the canvas in a flow for more complex data processing.

You can add modeling nodes to your flow. Each of the modeling nodes is a different modeling technique. You can add several modeling nodes to your flow to try different modeling techniques with your data. After you finish setting up your flow, you can run it so that your data is processed and the analyzed by the modeling nodes.

The modeling process

The ability to predict an outcome is the central goal of predictive analytics, and understanding the modeling process is the key to using SPSS Modeler.

A model is a set of rules, formulas, or equations that can be used to predict an outcome based on a set of input fields or variables from your data. For example, a financial institution might use a model to predict whether loan applicants are likely to be good or bad risks, based on information that is already known about past applicants. A tuned model is one of the objectives of working in SPSS Modeler.

Modeling is conducted in multiple iterations. Typically, you might run several modeling nodes that use the default parameters. You can then fine-tune the parameters for the modeling node, or you might return to the data preparation phase to adjust the data for the modeling node.

SPSS Modeler offers various modeling methods that are taken from machine learning, artificial intelligence, and statistics. You can use the methods available on the node palette to derive new information from your data and to develop predictive models. Each method has certain strengths and is best suited for particular types of problems.

If you get promising results from one of these modeling nodes, then you can save the modeling node as a model. This model can then be promoted and deployed for use in real-time predictive analysis.

Nodes palette

The nodes palette has all the nodes that are available in SPSS Modeler. They are organized into groups based on their function. You can add any node in the nodes palette to your flow.

For more information about the nodes palette, see Nodes palette.

SuperNodes

You can save all or part of your flow as a SuperNode. This SuperNode can then be used just like a node in other flows. You can use a SuperNode to add complex layers of processing to a flow without adding a long sequence of nodes that can clutter your canvas.

For more information about SuperNodes, see SuperNodes.