Partilhar via


Column Mapping Tab (Mining Accuracy Chart View)

Use the Column Mapping tab to provide information to Microsoft SQL Server 2005 Analysis Services (SSAS) that describes the data and the mappings that you want to use to test the accuracy of mining models.

For more information:Column Mappings (Lift Chart), Validating Data Mining Models

Options

  • Mining Structure
    Displays the selected mining structure that contains the models for which you want to test the accuracy.
  • Select Structure
    Opens the Select Mining Structure dialog box.
  • Select Input Table(s)
    Displays the selected input tables used to generate the lift chart. Use this to select the table that contains the test data that you will use to test the accuracy of the models.
  • Remove Table
    Removes the selected table. This button is disabled if a table has not been selected or does not exist.
  • Select Case Table
    Opens the Select Table dialog box. This button appears only if a case table has not been selected.
  • Select Nested Table
    Opens the Select Table dialog box. This button appears only if a case table has been selected. If the associated mining structure does not contain a nested table, this button is disabled.
  • Modify Join
    Opens the Specify Nested Join dialog box. This button is active only if the nested table is selected.
  • Filter the input data used to generate the lift chart
    A grid that contains columns that you can use to filter the input data. With this grid you are essentially creating a WHERE clause in the query that returns the data from the source data.

    Value Description

    Source

    Select the source table from the drop-down list.

    Field

    Select a column from a list of columns derived from the source table.

    Group

    Use with the And/Or column to group expressions together.

    And/Or

    Use to create a logical query.

    Criteria/Argument

    Specify a condition or user expression that applies to the column.

  • Synchronize Prediction Columns and Values
    Check to coordinate the predictable attributes in the grid so that, even if they have a different name, they are derived from the same predictable mining structure column during model training.

    Clearing this box is an advanced feature. You should only clear this box for cases in which you know that two mining structure columns derive from the same underlying relational or multi-dimensional source, and in which they have the same states or discretization.

  • Select predictable mining model columns to show in the lift chart
    A grid that contains columns to control which models are included in the lift chart and how they are used by the lift chart.

    Value Description

    Show

    Check to include the associated mining model in the lift chart.

    Mining Model

    Lists the mining models that are contained within the mining structure.

    Predictable Column Name

    Select a predictable column that is contained within the mining models that will be used to create the lift chart.

    Predict Value

    Select a value for the predicable column. If you leave this blank, the lift chart predicts how well the model performs for all states of the predictable column.

See Also

Reference

Mining Accuracy Chart View (Data Mining Model Designer)

Other Resources

Mining Accuracy Chart Tab How-to Topics

Help and Information

Getting SQL Server 2005 Assistance