Oracle Data Mining Examples
A number of sample streams are included that demonstrate the use of ODM with IBM® SPSS® Modeler. These streams can be found in the IBM SPSS Modeler installation folder under \Demos\Database_Modelling\Oracle Data Mining\.
Note: The Demos folder can be accessed from the IBM SPSS Modeler program group on the Windows Start menu.
The streams in the following table can be used together in sequence as an example of the database mining process, using the Support Vector Machine (SVM) algorithm that is provided with Oracle Data Mining:
| Stream | Description |
|---|---|
| 1_upload_data.str | Used to clean and upload data from a flat file into the database. |
| 2_explore_data.str | Provides an example of data exploration with IBM SPSS Modeler |
| 3_build_model.str | Builds the model using the database-native algorithm. |
| 4_evaluate_model.str | Used as an example of model evaluation with IBM SPSS Modeler |
| 5_deploy_model.str | Deploys the model for in-database scoring. |
Note: In order to run the example, streams must be executed in order. In addition, source and modeling nodes in each stream must be updated to reference a valid data source for the database you want to use.
The dataset used in the example streams concerns credit card applications and presents a classification problem with a mixture of categorical and continuous predictors. For more information about this dataset, see the crx.names file in the same folder as the sample streams.
This dataset is available from the UCI Machine Learning Repository at ftp://ftp.ics.uci.edu/pub/machine-learning-databases/credit-screening/.