Select the Case Table
The mining activity uses a table or view to build a model, or it applies a model to a table or view. See Case Tables for information about how case tables can be built.
For optimal performance of the wizard, the case table and any joined tables or views must be set up properly. See Setup for Tables for details.
Specify the table or view to be used by the mining activity:
- Schema: The schema containing the input table or view.
The wizard lists all schemas containing tables and views
that have been granted access to the user account that you are connected to. The schemas are listed in alphabetical order.
- Table or View: Table or view to use. Only tables that have been granted select rights to Oracle Data Mining are listed. If you invoked the wizard from a context menu, this field contains the name of the table or view that you selected.
- Join additional data with case table. This selection is displayed for build activities only. Check this box to join additional data, such as a text column, to the case table. The default is to not check the box. It is important to set up any additional tables properly; see Setup for Tables.
- Unique Identifier: If possible, specify an attribute, such as a customer ID, that uniquely identifies each case. The choices are:
- Single Key: The default is to specify a single attribute; select the attribute from the dropdown list.
- Compound or None: Select this choice if you need to use two or more attributes to uniquely identify a case or if there is no unique identifier. The wizard will create a supporting table. Creating this supporting table may require a significant amount of time and disk resources.
- Select Columns: You can include all usable columns, the default, or you can select only certain columns. To deselect a column, click the checkbox.
- Sampling Settings: The wizard calculates statistics for the data. To ensure good performance, all statistics are calculated on a sample of the data. For information about calculating statistics on a sample, see Calculating Statistics. It is important to have an appropriate sampling strategy and an appropriate sample size. Click Sampling Settings to display the default sampling settings for the case table. The type of sampling and the size of the sample are displayed; you can edit these settings.
Click Next to continue.
If you checked Join additional data with case tables, you now define the join.
The wizard validates that you have specified a table or a view and that the schema and table or view exist in the database. The wizard also calculates statistics for the data by generating a sample and then calculating statistics using that sample. If you selected Single Key, the wizard validates that you selected an attribute. If you selected Compound or None, the wizard creates the supporting table when you click Finish.
Copyright © 2006, 2008, Oracle. All rights
reserved.