User Tools

Site Tools


Using Filters

Filtering enables you to easily create models built on subsets of your data. The filter is applied only to the model and does not change the underlying data source.

Susan supposes that people from 46 and older may be a good target group for the marketing campaign.

First Susan makes a copy of the STM-Decision-Tree model.´

Create a new Model

  • Click the Mining Models tab.
  • Right click the STM-Decision-Tree model, and select New Mining Model.
  • In the Model name field, type STM-Decision-Tree-older-45.
  • Click OK.

Next, create a filter to select customers for the model based on their age.

  • Select the model. On the Mining Model menu, select Set Model Filter.
  • In the Model Filter dialog box, click the top row in the grid, in the Mining Structure Column text box.

The drop-down list displays only the names of the columns in that table.

  • In the Mining Structure Column text box, select Customer Age.

The icon at the left side of the text box changes to indicate that the selected item is a table or a column.

  • Click the Operator text box and select the greater (>) operator from the list.

:-A Decision Tree older than 45. cannot select < operator for customer age

  • Click the Value text box, and type 45.
  • Click the next row in the grid.
  • Click OK to close the Model Filter.

The filter displays in the Properties window. Alternately, you can launch the Model Filter dialog from the Properties window.

Build and Deploy your new Model

You now have two new models displayed in the Mining Models tab.

  • Right-click the STM-Decision-Tree-older-45 and select Process Mining Structure and all Models.
  • Click Run to process the new models.
  • After processing is complete, click Close on both processing window.

Test the lift of the filtered models

  • Switch to the Mining Accuracy Chart tab in Data Mining Designer and select the Input Selection tab.
  • In the Select data set to be used for Accuracy Chart group box, select Use mining structure test cases.
  • On the Input Selection tab of Data Mining Designer, under Select predictable mining model columns to show in the lift chart, select the checkbox for Synchronize Prediction Columns and Values.
  • In the Predictable Column Name column, verify that Bike Buyer is selected for each model.
  • In the Show column, select each of the models.
  • In the Predict Value column, select “yes”.

  • Select the Lift Chart tab to display the lift chart.
  • Move the Population percentage to 30%.

It is interesting to compare the models. The unfiltered Decision Tree model appears to capture more potential customers, but when you target customers with a prediction probability score of 76,95 percent, you also have a 24 percent chance of sending a mailing to someone who will not buy a bike. Therefore, if you were deciding which model is better, you would want to balance the greater precision and smaller target size of the filtered model against the selectiveness of the basic model.

As said, the value for Score helps you compare models by calculating the effectiveness of the model across a normalized population. A higher score is better, so in this case you might decide that targeting customers older than 45 years is the most effective strategy, despite the lower prediction probability, because the target population score is higher than the unfiltered Decision Tree model.

Enable Drill-Through on the Mining model

If a mining model has been configured to let you drill through to model cases, when you browse the model, you can retrieve detailed information about the cases that were used to create the model.

  • On the Mining Models tab of Data Mining Designer, right-click the STM-Decision-Tree-older-45, and select Properties.
  • In the Properties windows, click AllowDrillThrough, and select True.
  • In the Mining Models tab, right-click the model, and select Process Model.
  • Switch to the Mining Model Viewer, set the background for the Buke Buyer Flag to yes and drill thought the model and structure data.
bicn01/dm07.txt · Last modified: 2018/12/04 08:39 (external edit)