Overview of Binning Explorer

The Binning Explorer app enables you to interactively bin credit scorecard data. Use the Binning Explorer to:

• Select an automatic binning algorithm with an option to bin missing data. (For more information on algorithms for automatic binning, see autobinning.)

• Shift bin boundaries.

• Split bins.

• Merge bins.

• Export a creditscorecard object or generate a function to create a creditscorecard object.

Binning Explorer complements the overall workflow for developing a credit scorecard model. Use screenpredictors to pare down a potentially large set of predictors to a subset that is most predictive of the credit scorecard response variable. You can then use this subset of predictors when using the Binning Explorer to create the creditscorecard object.

Follow these steps to get started using the Binning Explorer app:

1. Open the Binning Explorer app.

   • MATLAB^® toolstrip: On the Apps tab, under Computational Finance, click the app icon.

   • MATLAB command prompt:

     • Enter binningExplorer to open the Binning Explorer app.

     • Enter binningExplorer(data) or binningExplorer(data,Name=Value) to open a table, data, in the Binning Explorer app by specifying optional name-value arguments. Use these arguments to set the properties of a creditscorecard object during a Binning Explorer session. For a list of arguments, see creditscorecard.

     • Enter binningExplorer(sc) to open a creditscorecard object in the Binning Explorer app by specifying a creditscorecard object (sc) as input.

2. Import the data into the app.

   You can import data into Binning Explorer by either starting directly from a data set or by loading an existing creditscorecard object from the MATLAB workspace.

3. Use Binning Explorer to work interactively with the binning assignments for a scorecard.

4. Export the scorecard to a new creditscorecard object or generate a function that creates a creditscorecard object.

You can follow these steps when using creditscorecard object functions in Financial Toolbox™:

1. Fit a logistic regression model.

2. Review and format the credit scorecard points.

3. Score the data.

4. Calculate the probabilities of default for the data.

5. Validate the quality of the credit scorecard model.

For more detailed information on this workflow, see Bin Data to Create Credit Scorecards Using Binning Explorer.